(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
See other formats

Full text of "Text Book Of Mechanical Engineering"



flftecbantcal Engineering





' M.INST.C.E., M.I.MBCH.E,, M.I.E. 1C.






.  .       '      ' 1912,   -.

AM. Jt/tf/fTS. /tfiSKRVtil*.


«!                     T T is now many years since the initiative of the City and

I,                 A     Guilds of London Institute, in providing an examlna™

f|'                 tion for Mechanical Engineers, first suggested to me the

desirability of writing the present text-book. In preparing
students for this examination, I was being constantly asked
for a comprehensive work which would at least show them
the general lines on which their study, as engineer
apprentices, should proceed ; and, in seeking to meet their
request, I had to consider seriously (i) whether the whole
theory and practice of Mechanical Engineering, or even a
pnfcis of It, could be compressed into one volume, and (2)
whether It was desirable so to compress it. That this work
has here been written is sufficient evidence of my-t>wn
solution of the above questions—a solution which has been
fully confirmed by the successful use, in teaching engineer
students, of my chapters during the years of their prepara-
tion. I am perfectly aware that there are many who will
object to any attempt to convey the rationale of practical
processes by description on paper; others may accuse me
of * cramming/ by attempting to condense the theories of
engineering into half a volume* 1 would earnestly ask all
these gentlemen, before condemning what may seem to
them a too ambitious undertaking, to first consider care-
fully the following reasons which appeared to rne to
support my decision :-—fi) The saving- of time to the
student, who need not now be always ' beginning at the


had i»f

j  «*f  'JM

MUI ja^
II may *i
^» v*« 4 4
^   thr r

t, ' \*



s        ifr *M

tluk (!is*tdvat)ta;it* **f having f*
small   u*xt-lxM>ktt ;   '*l sivin
initcir wiifii nm'ssary t«> JHV
needles-* ivjirtit'ifjft :           hw

in a              vohtim*, I'liibnicln;;

u'C<'^sitj' to bt* ////;  (

I wrifrf*    to %vit, KaiiKi

''.fl th«* £n I that j*r»i^lii4

rM'rilHnl*           jJMUlt' n* ^i

aiul In tin* C'ity <lutl»l'*'

I            »t*W                        fir^!«wl


I*             tin*          nl th

tif                 Iht?                                iti|^4»fi^.     Ti«

i»* l«                   flu*                        llir |

J)ff H,'t*W«*                   II             flf                                        ^

A                              I1* ii*iri%/f4 f^r   ?*|

lit tlir             im                                   I

in                                            ifil

I                                MI l|


ttf tic

tht                    1                       in          Ih" i

out                    «f *                         I aitt

i i# far                          but the

i»t          tiie           itiii            UN!

the                       the           to It*

tti            tttl I         will fat

In Part ft I                     I             «f all tht fit

ami                                          lay the

rnit him

dtic        fif                     to* ,tiE

Is            ttf                 em

i             t»t         «f

c»n                      it ; lit my

/i                                            >

Preface.                                   ix

welcomed which attempts to lay the demon of l rule of
thumb,' the autocrat of even my own apprentice days. To
encourage exactitude and prevent one source of error in
the application of formulae, I commence with a ' synopsis
of lettering,' and have here introduced what I believe to be
much needed, the retention of a certain letter wherever
possible solely for one purpose. Though this was not
always entirely practicable, I yet venture to think that
some improvement has been effected. It is unfortunate,
for example, that /stands both for stress and acceleration ;
but at           it need not be adopted both for tons and

pounds. I have, therefore, employed it for tons only*
Again, velocity per minute and per second are better
separately distinguished, as in the text, by the letters
V and v respectively.

While never introducing mathematics unnecessarily, I
have stated all the 'steps' that space permitted in such
mathematics as have been introduced, and the latter will
be found of but an elementary character, involving only
simple equations, fractions, and the use of tables of
and logarithms. The substitution of graphic treatment for
the higher mathematics in many cases will, I thinkr be
appreciated by most students.

As regards the order of Part II., the Strength of
Materials without doubt comes first, to be followed by
Energy and Kinematics; these all assist in the treatment
of Prime Movers worked by           or liquid*. With the

knowledge acquired frdm Fart L and his own                                 i

in the workshop, supplemented by the theory of Part IL,                |

the student should be able to commence the study of                £

original design, for he is now in, acquaintance          with                 1

what theory directs and the woricihop                                             \

Regarding itiuMmtlonft* 1 commenced with iht                             .     *

of admitting no' highly-shaded                               which*

showing nothing of Interior parti,' are only                  to

I        '   '


x                                     Preface.

confijse the student. Elaborate drawings, of course, ne-
cessitated great labour on rny part, as well as considerable
co-operation from makers and the editors of engineering
journals. Such aid has,in every case been afforded most
ungrudgingly, and in many cases has exceeded my most
sanguine hopes, both as regards, drawings and* matter.
The necessity of well-detailed, modern examples, has
always been present to me, and I confidently believe that
such have been supplied. In connection with these, I
would ask the reader to uriite with me in thanking the
following firms and gentlemen who have $o kindly
helped :—                         ,

Messrs, the Britannia Company, Colchester.
„    . George Booth & Co., Halifax.
„       Joshua Buckton & Co., Leeds.
v/    Mr. John Cochrane, Barrhead, N.'B.
1 Mons. Delamare-Deboutteville, Rouen.
Messrs, the East Ferry Road Engineering Works

Company, MillwalK
The Editors of Engifmriug', London.
Messrs. Greenwood & Batley, Leeds.
„       Andrew Handyside & Co., Derby.
„       Hulse & Co., Manchester.
„       B. & S. Massey, Manchester.
„       Priestman Bros., Hull
„       David Rollo & Sons, Liverpool,
„       Samuelson & Co., Banbury,
n      Selig^ Sonnent|ial & Co., London,
„       James Simpson & Co., Pimllox
n       Smith, Seacock, and Tannctt, Leeds.
„       Smith & Coventry, Manchester.
•>      .    „      the Sturtevant Blower Company, Boston,

U.S.A.t and London,
jj,      Tangyes Limited* Birmingham.
Mr. Ralph H. Tweddell^ Westminster.
'    Sir J. Whitworth & Co,,' Manchester.
Mr. Wilson Wof^del!,- of the N.E. Railway.



I have also to thank my assistants at the Goldsmiths'
Institute, Mr. William Ashton and Mr. George T. White,
for much valuable aid in the correction of proofs, and
Mr. R. W. Weekes for assistance in the matter of electric

' In conclusion, it is iny sincere wish that the? hook may
prove of real benefit to engineers of every da.ss. In
furtherance of this, I will gladly explain any portion that
may seem abstruse, and shall be greatly obliged by having
any errors pointed out. I nni.st finally .state that 1 do not
Intend tha work to be merdy an aid to any particular
examination, but I have introduced whatever seemed to
me most helpful to those: for whom it has been prepared.


(hhhmiths* Institute, New ^Vvw, A*»A\
October, r 894.


rPHE  demand for a  Second Edition of thin book etc*
•*•     curring so  soon,   the   Kirst   Kdition   having
exhausted within  a  few  month* of publication,  1
only had opportunity to  make the tmial
to add the more necessary                  in nn Appendix.

Speaking generally,  I   nm highly gratified at the ft*
ception which has been                 to tie workfc

considering the imperfection of it* l*lr«t  Edition! me!
here take occasion to sincerely thank                       and

reviewers both for praise and far help*

As there has been            little miiunder^ttttctitig in thc%

reading of my first preface* 1st me at
only examination 1 had In view

that which is the                              for the

* his actual requirements in           the            work/

Once more 1

ge«tion§f renewing at the                    my         to

any difficult portion.                    WjurM|>

tioldmttkf iMtthtt**

 Cri* w £.
	?y H
 C ft
	s 3 m
 :     "^
	•Jt «-»• 0 o
 sr C '
 n £
	— 1 5^ ^t»

•** ^^
	o ft

n H *
 C  P
 r^ p*
	"A JJ8

	5* n
	f5 55*

^   *. *n   *
	- (
	*  *

I II 3  sl = S"-^"! ~ If S-9   ~

~  ~ 2 ^s'lic-IH.s * i =

«t   •*?              ****


Preface.                                   xiii

carefully re-read and corrected. The excellent and
ready method of treating redundant members given on
pp. 924-5 is due to Mr. Max am Ende, though obtained
from Prof. Pullen's work on Graphic Methods', and here
further simplified. Other sources of information have
been acknowledged in the text.
/-* 7j vz > T y/ /                     WILFRID J LINEHAM.

Goldsmiths' Institute,                                           J

January, 1900.________________


A MOTHER short Appendix has been added to bring
-^ the work up to date, and the general text has been
thoroughly revised.                     WILFRID J. LlNEHAM.

Goldsmiths* Institute,
October, 1901


HPHE continued progress of the Science of Mechanical
-*- Engineering necessitates a further Appendix, the
fifth ; for a book of this kind must be kept up to date
at all costs. We have been requested to publish the
Appendices separately, but it is found impossible to do
this without loss * so are constrained to advise readers to
transfer their early editions to younger students, for whom
they are still useful

We have sometimes been asked to divide the book
into separate volumes, but are just as often urged to keep
it intact. In view of these conflicting requests, it appears
advisable to retain the present form of a self-contained
manual, the want of which was the original reason for
writing the book.

It is impossible to specialise in $uch a work.
matter can only be taken as a general statement of scietide
and practice; and types only can be used as examples.
I trust that specialists will consider this point whenever
they are apt to criticise a little harphly the chosen typ^s.

Lastly, I would urge readers to mtfee regular use of the .
Appendices, which are fully connected by reference notes v




with  the body of the  book ;  and  I   have on<
express my gratitude to correspondents and c

for their continued help.            ,tr             T   T

^            Wru-'RiD J. L

Goldsmiths' Institute,

January, 1903.______________


T   HAVE to thank the following firms for cc
•*•    to the Sixth and Seventh Editions:—
Messrs.* Bopp & Reuther, Mannheim.

Greenwood & Batley, Leeds.

Palfreyman, Liverpool.

Alfred Herbert, Limited, Coventry.

Pfeil & Co., London (acting for J. Reinec

Fletcher, Russell, & Co.

Waygood & Otis, London.

^ /^   v/ , /  ,v *                     WILFRID J. L

Goldsmiths' Institute,                                           J

March, 1904.


TTHIS Edition has been considerably enlarge
•*•     to still keep the book fully abreast of
and the following firms and gentlemen have rj
assisted me with matter for Appendix VI. :—

Messrs, Samuelson & Co. Banbury, Pfeil & Cj
Alfred Herbert, Ltd., Coventry, Herbert Morrid
and Crossley Bros. Messrs, Andrew Forster, (j
ball, J. Hartley Wicksteed, R K. Kennard,'
W. W. F. Fallen ; to all of whom and to othc
te'nder my sincere thanks.           ^             | j

1 CoUtgt*

ISfmember, 1905.

%                     PREFACE TO  THE

ENERAL revision          correction only





C ASTI NO AN I) M O U 1,1 )I Nt 1

Varieties of Cast Iron
The Cupola
Moulding Sand.
Methods of Moulding
Examples in Greensand
Examples in Loam   .
Machine Moulding  .
Chilled Castings
Malleable Castings   .
Brass Founding
Mixtures   .










Wood used       .........    43

Pattern Building........49

Core Boxes and Prints.......55

Wheel Patterns........5«

Striking Boards        .       .        ,.....61.

Contraction Allowance     .......    62

Plate Moulding', Stopping-ofF, &c.    .       .       .       .       .64

Crystallisation ancl Unequal Shrinkage   *       .       .       .67




Chemical Elements.......

Cast and Wrought Iron......


Bronzes and Brasses.......





The Hearth......



Smaller Steam Hammer.

Heating and Welding    ....

Examples of Simple Forging .

Examples of Heavy Forging and Stamping



The Forge......

Heavy Steam Hammer ....

Piled-up or Scrap Forgings    .

Heavy Steel Forging      ....




Classification ......

General Principles .....

Tool Angles   ..,,.„

The Screw-cutting Lathe

Supporting and Driving Lathe Work    *

Chucks and Toolholders        ,      *

'The Break Lathe   .

The Boring Machine                           *

Drilling Machines         *

The Planing Machine    -

The Shaping Machine   ,

The Slotting Machine   .





The Milling Machine
Milling- Cutters
Machine Vice .



The Marker-ofPs Tools.......    183

The Fitter's Tools........    186

Machinist's Requirements......    195

Capstan Lathe........    200

Erector's Tools........    202

General Processes........    209

Application to the parts of a Horizontal Engine    .       .    215
Regulator Gear        .       .        .        .        .        .        ,216

Valve Rods........    226

Eccentrics        ........    230

Slide Bars and Brackets......    232

Crank Shaft Bearings......    235

Crank Shaft........    238

Connecting Rod      .......    240

Crosshead and Piston     ......    245

Governor Gear        .......    249

Cylinder and Valves     •......    257

Fly-wheel and Bed Plate......    262

Brass Work........    264

Erecting the Engine.....•  .    268

Sundry Notes.........    274


Materials........»    279

Hand Tools and Hand Processes.....    283

Punching v. Drilling.......    287

Punching and Shearing Machines .....    289

Plate-edge Planing Machine ......    294

Bending Rolls........    297

Flanging Presses........    300

Drilling Machines ........    303

Hydraulic Riveting Machines        .               .       ,       .    313



Locomotive Boiler-shop .......

Marine Boiler-shop......

Ship Yard.........

Pneumatic Caulker and other Tools       ....

Electric Welding........

Description of Boilers: Lancashire, Marine, Locomotive,

Tubulous, and Vertical......

Geometry        .  '   -.

Setting-out a Marine Boiler......

Riveting the Boiler........

Setting-out other Boilers......


Synopsis of Lettering



Stress, Strain, and Elasticity.....

Work Diagram........

Stress due to Impulsive Load, and to Heat   .

Testing Machines........


Shackles for Specimens.......

Strain Measuring   ........

Stress-strain Diagrams   .......

Wohler's Law and Factor of Safety       ....

Table of Stresses for Different Materials       .       .       «

Classification of Stress-action.....

Tension Stress-action:—

Ropes, Pipes and Cylinders, Flywheel, Bolts, &c. ,
Compressive Stress-action . . . . . i
Shear Stress-action :—

Suspension   Link,   Riveted  Joints,   Cotter   Joint
Shafts, Coupling Bolts, Keys, Springs   .


Bending Stress-action and Theory of Beams:—    .

Neutral Axis and Moment of Resistance

Bending Moment and Shear   ....

Theorem of Three Moments   ....

Examples of Beams.....

Deflection of Beams.....

Combined Bending and Tension Stress-action
Combined Bending and Compressive Stress-action

Pillars and Struts......

Furnace Tubes......

Combined Torsion and Bending    ....

Combined Torsion and Compression

Framed Structures.......





Force, Mass, Velocity, and Momentum.

Energy Forms : Conservation, and Transformation

Transmitters of Power • .

Simple Machines    .        .               .       .

Kinematics: Lower Pairing   .               ...

The Slider Crank Chain .        .       .     •  .

The Quadric Crank Chain

Higher Pairing......

Velocity and Acceleration Curves ....

Link Work......'       .       .

Shafting, Bearings, &c.......

Spur Gearing........

Bevel Gearing........

Worm and Screw Gearing.....

Epicyclic Trains     .        .        .

Belt Gearing  .       .......

Cotton-rope Gearing       .       .        .

Wire-rope Gearing.......

Pitch-chain Gearing......

Compressed-air Transmission         ....

Hydraulic Transmission......

Electric Transmission......

Friction and Work Lost......

Friction Gearing.......

"4   ,*-'"





r f.

I 'ynavnometers

Kitkiencies of

ON H' \:

Kx< .INFS.

1 Hii.imical Theory of ! !e,it     .,        .         .         .         -        ,    ;8i
Transfer of Heat     .        .         ,        .         ,.         ,        .        .581
Measurement of Heat      .        .        .         .         ...    584

Expansion of Gases         ...         .         ...    587

Latent Heat    .....         .         .              .    591

Saturated and Superheated Steam.         .         .              .    595

Mechanical Equivalent of Heat      .....    599

Internal and External Work  .               .         ...    600

Specific Heats of a Gas .                      .         .                  602

Isothermals and Adiabatics    .               .         ...    605

Carnot's Engine and Reversible Cycle   ....    608

Losses in Steam Engines        .       .        .         .        .        .613

Expansion in Cylinder    .        .       .        ,         .        .        .615
The Indicator and Indicator Diagrams .         .        .        .616

Multiple-stage Expansion       .        .         .         .        .        .621

Combination of Indicator Cards    .....    622

General Idea of various Steam Engines.         .        .        .    627

Distribution of Steam: by Cataract        ....    63.4

„              „            by Eccentric      ....    636

„              „            by Link Motion.        .        .       .   640

„              „            by Radial Gear....    642

Governors        .....         .         .         647 & 655

Variable Expansion-Gear       ......   650

Automatic Expansion-Gear    .               .         ...    654

Trip Gears       ......         ...    656

Zeuner's Valve Diagram....         ...   660

Ideal Indicator Diagrams for Compound Engines.        .   666
Correction of Indicator Diagram for Inertia .        .        .   673
Curves of Crank Effort   .......   676

Weight of Fly-wheel       .        .        ..        .               .           679

Horizontal Compound Engine        .....    681

Triple Expansion Marine Engine .....   685

Condensers     .....        .         ...   686

Marine Details        ........   688

Compound Locomotive  ....         ...    689


Tractive Force

Boiler Fittings

Combustion and Forced Draught

The Gas Engine

Petroleum Engines






Head, Pressure, and Velocity Energy    .        .        .
The Jet Pump ......        .        .

Discharge of Water from Orifices .


Measurement of Stream Horse-Power by Gauge Notches    7 14

Friction in Pipes, and Virtual Slope       .        .        .        -    7J5

Loss by Shock        .       .       ......    716

Principle of Momentum applied to Water Wheels .       -7*7

Water Wheels         ........    720

Turbines ..........    723

The Centrifugal Pump   .......    728

The Impulse Ram .       .    .   ......    729

Piston Pumps .........    73°

The Pulsometer      ........    735

The Hydraulic Press      .       ......    735

The Accumulator and Hydraulic Transmission     .       .    736

Hydraulic Lifts : and Intensifies   .....    73$

Hydraulic Cranes   .       .       .       .       .       .       .       .    740

Hydraulic Pressure Engines  ......    742

APPENDIX I. (to Second Edition) .... 747
APPENDIX II. (to Third Edition) .... 779
APPENDIX III. (to Fourth Edition) . . . .917

APPENDIX IV. (to Fifth Edition) .....   949


APPENDIX V, (to Sixth Edition)  ......   968

APPENDIX VI. (to Eighth Edition)     .       .       .       . ion

LIST  < U«   1M.ATKS.


	Arrangement of , in Kn^i'ieer'^ Mn;tln       -        •   ."..*. v loiut. Steam Hammer    .                           ...   ,-,..-( Steel-tempering Diagram          ....   fr:e ^ tons Steam Hammer     .....   .',.v<f
	iyC   1 •;8
 126 ! ^O

	so ins. Centres Screw-cutting (iap Lathe         .  f,\e Treble -geared, Screw-cutting Break Lathe     .   f,u-e Boring Machine and Engine Combined .        .   t\n-e Double-geared Drilling Machine     .        .       . >v^ Planing Machine     ...... fine
	•* J 150 1 60 162 I64 I 7O -

	12 ins. Stroke Shaping Machine      .        .        .   A.Y<- .v Slotting Machine      ....        .           face Universal Milling Machine      ....   face Drilling Machine for Marine Boiler Shells     .   Az(i- is Multiple Drilling Machine       .... face Hydraulic   Machine   Tools   for   Locomotive Boiler Work         ...... f^ce   ' Hydraulic Machine Tools for- .Marine Boiler Work           .        .        .                  .         .        .   face
	1   /  V
 174 1 80 310-11 312
 318 -« *>o

	Triple-expansion Marine Engines  .        . . .   .   /;/« Compound Express Locomotive     .        .       .  face 'Hexagon7 Lathe   .       .        .        .        ... face
 692   '    .
 986 ;






UP to the time of Watt, and even later, a very great deal of
wood was used in engineering structures, even to the extent of
steam pipes, but as fluid pressures became higher, other materials
were sought, and cast iron was the first to recommend itself.

Cast Iron is the most crude form of the metal, and is
obtained direct from the blast furnace by the fusing of the ore
with some flux, which varies according to the
nature of the particular ore, sometimes requir-
ing clay, but in this country usually lime. The
molten iron runs down into channels or pigs
and is then called pig iron, while the slag is withdrawn

Of the pig iron thus formed there are eight commercial
varieties, according to the quality of the ore and the blast used;
thus, increase of blast and diminution of fuel gives a whiter iron.

- 81

J   7 I White   (silvery,    hard,   and    strong),   for   conversion
£   6 [         into wrought iron.

I sj

*u  4    Mottled.    Strong foundry iron.

1   3)

1   2 > Grey (soft and weak) for ornamental castings.

T )

2                          Composition of Cast Iron.

Most of the impurities disappear in the blast furnace, but
carbon is absorbed from the coke fuel, and the presence of this
carbon, mechanically mixed in the form of graphite, makes the
iron more liquid when molten, but at the same time produces
weakness in the casting. ,'Th.ere is never more than five per cent,
of uncornbined carbon, while in the white iron there is almost
none, it being chemically combined, and then actually increases
the strength of the iron.

Table showing chemical composition of the three principal
varieties of pig iron, in percentages :—

Iron  ...................
	Grey. 90*24.     ....
	Me 8c
 )*3 79 *ii
 "17 •48 •17   . •6
	White. 89-86 2-46 •87
 2-52 •91

Carbon (combined)         ...
	'I'O2      .....
 2*64   ..... vo6    .....
Graphite (uncornbined) ..... Silicon ................ . ........

	•>      .   ••••: 1*14.     ....
Manganese ....................
	:       -81   .....



^ The Cupola.—The pig iron is re-melted in the foundry in a
kind of small blast furnace called a Cupola. The cupola is re-lit
€very day (and is therefore not so economical as a blast furnace,
where the fire is never allowed to die out),* but this cannot be
avoided on account of the intermittent demand made upon it.

Fig. i is such a cupola, where the pigs and coke are raised by
the lift H L, hydraulic or otherwise, together with the man, who,
after breaking each pig in three, puts them all in at the door D,
charging as follows:—First, 7 cwts. of coke, next i ton of iron;
then^ alternately, 2 cwts. of coke and i ton of iron, until the
cupola is filled to D. The blast enters at B, and the mouth M is
stopped with luting clay. When all -the iror* is meked M is tapped,
and the metal taken away in ladks to the moulds.

During re-melting the iron is again apt to absorb impurities
from the fuel, such as oxides and silicates, the latter especially
producing more brittle material, and rendering the iron cold-short,
that is, easily snapped when cold. • Formerly, re-melting was
believed to be an improvement, and founders were advised to

* One blast furnace in the North of England, known to the writer, burned
for over twelve years incessantly, and was then only blown out for repairs.

The Cupola,


melt again and again,
even to twelve or thir-
teen times, but this
has now been demon-
strated to be a fallacy
(Seep. 779 ##^i o * i.)

To obtain a very
tough casting, such
as an hydraulic cy-
linder, wrought iron
turnings are some-
times mixed with the
pig in the cupola.

will now consider
how the moulder
forms his casting into
any desired shape.
To do this it is ne-
cessary in most cases
to make a wooden
pattern which shall
be the counterpart of
the casting required;
for several reasons we
shall see, however,
that the pattern will
not always be exactly
similar to the casting.
But more of this as
we advance.

The pattern is
impressed in sand
contained in two
moulding boxes, or
flasks, about half the
pattern in one box,
and half in the other;

Various Methods of Moulding.


JVlD.uLcLuiM JB^ooc^s.
COM MOM rwo-pA/ir BOX          Ifytf' 2,-

these'are sketched at Fig. 2. The boxes are light castings, ribbed
across as shewn, allowing space for the escape of gas from the
molten metal. (See Appendix //.,/. 781.)

Sand used in moulding is of two kinds, green sand and loam
^ Green sand is obtained from the chalk or coal measures, that
of the London basin being among the best. Green sand should
contain a large percentage of silica to give porosity, together with
a very little magnesia and alumina for binding purpose's. The
lining of Bessemer converters has about 85 percent, of silica in its
composition, while many moulders prefer to have as much as 93 or
96 per cent, of silica, leaving only 4 or 7 per cent, of other substances.
The sand should not burn on setting, or it will stick too much when
wetted for use again, and, while cohesion is necessary, it should at
the same time be porous enough to allow for the passage of air,
though not so much as to permit of any molten metal entering it.
v Loam is a mixture of clay (ferruginous or calcareous*) with a
considerable amount of rock sand (abraded rock). It is ground
in a mortar-mill and mixed with powdered charcoal, horse dung,
cow hair, chaff. &c., to give it binding power and porosity.

Besides the above, Cores require a mixture of rock sand and
sea sand (the latter for porosity), and Parting Sand, for the use
Implied by its name, consists of finely powdered blast-furnace
cinder, brickdust, or fine dust from castings; ail perfectly dry.

Moulding is practised by three different methods: Green
Sand, Dry Sand, and Loam Moulding.

* With iron or lime respectively. •

Various Methods of Moulding.


:ings, ribbed

;as from the

.d and loamj
easures, that
sand should
ogether with
poses. The
>f silica in its
mch as 93 or
:r substances.
) much when
, it should at
.ssage of airr
[ entering it.
sous*) with a
It is ground
, horse dung,

Dck sand and
i, for the use
Derfectly dry.
hods: Green

In green and dry sand moulding, patterns are generally used;
but in loam  moulding, which is only employed for objects of
regular form, the mould is struck out by means of a template, \
and built up by the moulder himself

* Green Sand is the geological name of a sand of very fine
texture. It appears black in the foundry because it is mixed with
a proportion of coal and charcoal dust; it is damped each time
that it is used. This is the most general method of moulding,
with castings not likely to warp too much by the more rapid
cooling. (See Appendix //.,/. 779.)

v Dry Sand is a mixture of old loam with an addition of
rock sand. It is so called because, after the pattern is moulded,
the sand is dried by means of fires hung in pans or trays over
the moulds. It is firmer and more suitable for the support of
long castings, such as pipes, columns, and large fly-wheels than
green sand is, and will produce finer castings, with less fear
of pieces of sand being torn away by the
flow of the metal. If pipes are moulded
in green sand, the tendency is to uneven
thickness in the castings, through sagging
of the sand.

• Loam Moulding, as we have said,
does not require a pattern, the mould
being struck in the pasty loam (the latter
being mixed with water) by means of a
rotating or sliding template, called a
* striking-board. Thus the core of a large
cylinder, is built up in brickwork, and then
covered with a layer of loam, which is
smoothed by a rotating striking - board

(see Fig. 3), much as a plasterer would work the cornice of a
house ceiling. Cubical moulds, such as those for condensers,
may also be worked in loam.

The simplest moulding done in green sand is called Open
Sand Moulding, and consists in laying the pattern in the sand
on the foundry floor, withdrawing, and then pouring in the metal,
a cover not being used. This is the method employed for such
common objects as moulding boxes (see Fig. 4).


6                               Simple Moulding.


A Cattle Trough.—Our next example of moi
ordinary cattle trough, and here two boxes are used
sand. The pattern may be of wood in the first inst;
the same shape as the finished casting. It is pi
bottom box, as in Figs. 5 and 6, and sand is filled in to
which is the parting. This parting is smoothed off, :
parting sand applied, and, the top box being put on 2
down, the whole is filled with sand and rammed w<
The top box must now be removed and the pattern
a slight rapping being given to effect its detachme
/ "sand, while the latter is dusted with Blackening, w
charcoal dust. But first, to make the blackening ad
meal is sprinkled on the mould, absorbing the damp
and thus becoming a pasty layer. The object of th*
is this: If the metal were to touch the side of the mo
enter into the sand surface, and thus produce a roi
This is allowable in moulding boxes, where roughness
advantage, but where a smooth casting is desired, b'
needed, as it ignites on being touched by the metal, a
a film of gas between it and the mould, a clean castir
result. (See pp. ^747 and 1012.)

Gates.—The mould having been sleeked and f
any little break in the sand mended, the gates havi
made for the entrance of the metal. Tapering plugs
usually left in the sand for that purpose, and these are n<

The more shallow the casting, the more gates a
many even as four.

Vent Holes are made in the more solid parts
(but not to touch the surface of the mould) to fi
passage of air from the latter.

The moulders, being provided with molten iron,

Gates and Vents.

ilding is an.

to hold the
ince, and of
aced in the
the line P P,
i dusting of
md fastened
*11 together.
taken away,
nt from the
rhich is oak
here, pease-
of the sand

1 blackening
uld it would
jgh casting,
is a decided
lackening is
.nd so forms
ig being the

inished, and

2  now to be
of wood are
DW removed.
re used ; as

of the sand
icilitate the

taken from

the cupola in ladles, as already described, pour it simultaneously
into the  gates   of the mould,  and the sand  being   afterwards

broken away, reveals the cast-
ing, which has filled the matrix
left by the pattern.

As regards the proper posi-
tion in which to lay the pattern,
a little thought is necessary, but
as a general rule the most un-
important part of the casting
should be upward, that being
the part to which the scum and
impurities rise. If possible, the
scum should be entirely re-
moved from the mould itself,
being allowed to fill a large
gate or projection. This is
especially done in the case of
steam cylinders, where purity
is a necessity. Gates should be
as central as possible, and have
their mouths a little higher than
the mould, but they should, as
a rule, enter the latter low down,
particularly in deep castings, in
order that the air may be made
to pass out at the vent holes ;
but much judgment has to be
exercised; and in most cases
they should be placed a little on
one side, namely, not to enter
on the top" surface, otherwise
the corners of the sand may be
knocked off by the force of the
flow; and finally, they should
be put where shrinkage is likely to occur, that they may tend to
. fill up any shrinking portion.           :

Our next example shall be a Hand Wheel for a large stop

Core Prints.

valve. The pattern here will be of the same shape as the finished
casting, excepting the square holes in the centre, which we will
suppose to be cast in, to be afterwards dressed up with a rough
file. Square core prints of any convenient length are put on either
side of the boxes, and of a diameter equal to the hole to be cast,
allowing a small amount for cleaning up (see pattern, Fig. 7).

A core box is now to be made, which is shown in Fig. 8, and
consists of two blocks of wood, hollowed out in such a way as to
represent the square hole required, and of a length equal to that
across the core prints from end to end The pattern is next
placed in the sand, as drawn in Fig. 9, and parting made, then
rammed up, with gate at fe, and vents here and there. Core sand
mixed with water is put in the core box (which fits together by
means of the pegs PP), smoothed off, removed, and placed in the
core stove to dry. The mould is finished and blackened, and
the core treated with black wash, which is charcoal dust mixed
with clay water, and used for the same purpose as blackening.
The core being put in place, as shown in Fig. 9, everything is
ready for the reception of the metal.

We shall now take the moulding of a Chain Pulley, by a
very ingenious method. Fig. 10 represents the pattern, with core

Pulley Moulding.


CJ^cujis JPsuZL&y.


prints for the centre hole, and divided in halves by a horizontal
plane. In all other respects it is the counterpart of the casting.
The operation is as follows: referring to Fig. n, .we must first
lay the bottom half of the pattern in the bottom box, and make
the parting b b ; next put in the top half of pattern, and make the
parting / /; lastly, fill up the box, and ram well together. The
pattern has now to be drawn out, and this is done by first lifting
off top box, taking away the top half of pattern, and returning top
box. Now, on turning the whole upside down, the bottom half
of pattern becomes the top, and may be similarly released. The
ring of sand M, it will be noticed, has all this time remained
resting on that half which happened to be at the bottom. It
is only necessary to make the core as in last example, put it in


Worm Wheel Moulding.

place by removing top box, form gates and vents, and complete

A Worm Wheel may have a pattern made in halves, and be
moulded in an exactly similar manner (see Fig. 12); the teeth on

the pattern being formed so as to gear with a wrought iron
worm which has been previously turned, the worm and wheel
pattern being rigged up on two axles to imitate their condition
when in actual work. In withdrawing the pattern from the sand
a slight screw motion must be given to allow for the angle of the

Moulding boxes are entirely or to some extent dispensed with,
and the floor of the foundry used for the reception of the pattern
wherever convenient; and then, except in such cases as that
shewn in Fig. 4, a cope or slab of sand is used, contained in a
box, to cover the impression. Examples of this kind of moulding,
with more or less complicated copes, will now be treated.

Fig. 13 is the plan of a Drilling Machine Table. The
pattern is of the same shape, with the exception of core prints
necessary for the slot holes. A core box is required for these
holes, and the whole is moulded face down, an extra piece being
ijeft in the casting at top, if thought necessary, to allow for scum.
Instead of the bottom box the floor might have been used, if
previously well vented with coke.

Perhaps the most ready way to mould the Cylinder Cover
shewn in Figs, 15 and 16 is to use three boxes (or what really comes



una 7>/////>.

Casting On.



to the same thing, twohoxes and the foundry floor), and make the
flange A loose on the pattern. On taking off the top box, this
flange maybe withdrawn, while on replacing the top box, and lifting
it and the bottom box together, the main pattern may then be
removed. The core for the centre is inserted in the usual way,
and the casting made in the position shown. The stakes s s fix the
position of the boxes with regard to the floor. (See App.II.,p. 781.)
v Casting" on.—Sometimes it is necessary to attach cast iron
to wrought iron in the mould itself, and so do away with the
expense of bolts. Casting on is the term adopted for the opera-
tion resorted to.

As an example, we will take the traction engine Road
"Wheel, shown in section, Fig. 17.

A core box is made as in Fig. 18, consisting of a slab of
wood A, with the boss B fastened to it, and of a hollow cylinder
of wood c to contain the core sand. Two cores are thus formed
and baked in the stove. A second core box is required, shown
in Fig. 19, consisting, as before, of a hollow box to hold the core,
and of the bosses DD in two parts, to make the impression for
the central part of the wheel nave. As the line oc in Jig. 18
corresponds to line x on Fig. 19, it will be seen that the prints
ppin Fig. 18 will leave spaces for the reception of the spokes.

It is only necessary to fix the spokes loosely in place by bolts to
wheel-rim, at the same time laying them in the spaces left for them
in the cores; build up according to
Fig. 20; add a central core, E, made
in ordinary core box; make gates, and
cast. The spokes are afterwards
riveted on wheel rim, and have the shape shewn in Fig. 21.




Loam Patterns.

Loam Moulding.—We shall now proceed to consider the
moulding of such objects as may be done wholly or partly in
loam by striking (or strickling), and first we will take an ordinary
Gas Pipe Main, with spigot and faucet, the former being the
smaller end of the pipe, which fits loosely into the faucet or larger
end of the next pipe, see Fig. 22, which represents the finished
pipe in section. To mould the outer envelope, we may either

r                                    CAST/ KG                                                                               *;


have a wooden pattern with the core prints at the ends, or may
strike out a loam pattern from a board. Assuming the latter
method, we need first a pair of trestles, Fig. 23, on which is
placed a hollow cast iron cylinder with journals at the ends, and
pierced with holes along its length for the venting of the core.

Round this cylinder straw rope is tightly coiled, after which a
layer of loam is laid on. The loam being dry, a second coating
is applied, and this time, as the handle is turned, the shape of
the core is struck out by means of a board B secured to the trestles.
The core b being dried in the stove, is blackwashed, and then
covered with another layer of loam to be struck out by the second
board A, and so the loam pattern is formed. Being again dried,
an impression is made in the mould, after which, the last applied
loam, or thickness piece, is removed, the blackwasb facilitating this,
and the internal core b being returned to its proper place in the
mould (see Fig. 24), gate is made, and casting performed as usual.

It is advisable to cast these pipes either vertically, or on an
incline, so that the metal may flow more easily and bring the
scum to. the end, and if they are very long, dry sand should be
used in the boxes instead of green sand, for reasons previously
stated. After the metal is poured, the escaping gas is lit at
either end of the pipe. (See Appendix //, p. 782.)

Fig. 25 represents the moulding of a Bend for the pipe in last


Bend in Loam.

example. To mould it we may either

have a complete pattern and core

box, when it would be done in

green sand, and  needs  no

further explanation; or it

may be worked in loam

by the aid of tem-
plates.    For the

latter method we

may  proceed   in

the following manner:
Take an iron plate,

Fig. 25, and on it fasten
the bent wire A of square
section as a guide for the
template B. Loam being laid,
it is traced out by B, which
gives it the form of the internal
surface. The length of the bend is
carefully measured off at c and D,
care being taken to allow an extra
piece at D to support the core, while at c
is applied the core box, E, giving the shape
of the internal faucet; the core is then care-
fully dried and black washed. Now apply more
loam and trace out by means of the larger
template G, which gives the necessary thickness,
while the faucet is supplied by the impression
of the core box H, and the spigot by that of j. The added
thickness is sbown by the dotted lines. Drying and black-
washing are again performed. Lastly, cover the whole plate
with loam, which, as it is required to be lifted off entire, must
be well stayed with cross and longitudinal wires tied together
by small wire. A wire should also stiffen the internal cores.
(The whole of the foregoing is repeated opposite hand, on the
plate F, the same bent wire being used.)

The solid block of loam thus formed on either board is
taken off, dried and blackwashed, and now all is ready for putting


Cf. 25.


Another Method for Bend.

together to form the mould, with the exception of the tl
layer on the internal core, which must first be removed, ;
two half cores taken off the plate and put back to bac
forming a complete central core. The whole mould is
at Fig. 26, the bent part of the core being supported by a
given in detail at K ; being a bent, tinned, plate.

It must be quite understood that if many castings are r<
the above operation would not be performed, as a patten
give more expeditious results.


A slightly different, but more usual way of moulding
bend, is to cast two plates, as at L, Fig. 2 6A, from a
pattern of the shape of the pipe, a little margin being left:
side, the figure being for a pipe having flanges at the end!
template takes the form M, and has a few nails driven in

Large Steam Cylinder.


prevent considerable wear as it travels along the plate. The
internal core is shown struck on the plate, and the opposite hand
having been made, these two are dried and laid aside. A larger



! !

template strikes out the patterns, right and left hand (the wooden
discs N serving as flanges), and these being dried, it is clear that
we now have two half loam patterns and two half cores; the
moulding, therefore, may take place in green sand in the usual

Steam Cylinder in Loam.                          21

way. A few sprigs or brads in the flanges N serve to fasten the
latter to the pattern.

The mould for a Large Steam Cylinder is usually made
entirely in loam, and this operation we will now examine.

Fig. 27 represents the casting in longitudinal section, elevation,
and in plan. The valve box is made separately, as is sometimes
done with these large cylinders, but in any case, no further
explanation is needed be^nd that previously given, as a pattern
would be made for it. The body of the cylinder would be swept
out entirely by template boards, but special projections, such as
steam ports and exhaust flange, would require core boxes and

An iron plate A, Fig 27^, is laid on the foundry floor to
support the structure, and a centre B is sunk beneath the ground-
line, an upright spindle c being taken of sufficient length, and
supported at the top by means of an arm D standing out either
from the wall or from a crane pillar; all is now ready to

A base of loam is swept out by the board E, shown in dotted
lines, and representing the bottom of the cylinder flange ; this is
dried and blackwashed, a flat ring d being then laid as a foundation
for the core structure.

Taking board E away, another (e) is used to strike out the
lower cylinder flange f, which is necessary as a support to help
plate d.

The loam/being dried and blackwashed, the external core of
the cylinder is next formed, because it is necessary to remove it
for the formation of the internal core, and the latter, being in one
piece and cumbrous, is made separately. The board F is now
used to strike the outer form, the central projection being for the
exhaust port, and an opening must be allowed at G, the full length
of the cylinder (see Fig. 29) for the reception of the port cores on
one side, and which may be traced out by a template board, while
a similar opening g, of the depth K j, must be left on the opposite
side for the exhaust flange core. It will be noticed that this outer
mould requires, for building, the aid of annular plates at H j K L M,
for the support of different pieces of the structure. These plates
do not go entirely round, being prevented by the ports at G, and

co/?£ aoxcs

Steam Cylinder in Loam.                        23

they now enable us to remove this outer portion in separate pieces-
to a safe place to dry, and allow us also to build up the internal
core. Thus plate K may be lifted by crane, removing the upper
portion first; next j and d. Discarding the loam plate f, which
is no longer needed, our next proceeding is to take the board N,
Fig. 28, having built up the core loosely with bricks, vented at the
joints with coke powder, and strike out loam to represent the in-
ternal surface of the cylinder j this is dried in place by open fires
and blackwashed. The projecting portions only now remain,,
which, as we have said, must be made from core boxes. Fig. 30
represents the box for the outer contour of the steam ports, and a
core is formed by laying it on a flat plate and filling up with loam.
The parts a a, of the core box should be noticed : sides b b, can
be easily taken away, but in order to draw away the centre c, the
flanges must be dovetailed to c in such a manner that they may
be left behind on withdrawal of the box.

This may be understood on reference to a, which shews one
of these loose pieces. They may afterwards be taken away in the
direction of the arrows. The box and core for the steam ports-
are shown in Fig. 31, and need no explanation.

The inside core for exhaust port, being circular, is struck out .
on a separate plate by board (Fig. 32), box p being required to
give the projection on the steam side, and Q for that on the
exhaust side. There is left the exhaust flange, which may be
formed from the box in Fig. 33, the flange itself being loose on
the pattern to enable the core to be withdrawn, the latter being
made on a plate similarly to Fig. 30.

s s are patterns for the web at top and bottom of the cylinder,
which, having been built into the core at Fig. 270:, may now be

Finally, all may be put together to form the mould, in the
manner drawn in Fig. 34, beginning at the bottom and putting
the different cores in their places as we proceed; chaplets being
required to support the annular exhaust core. Gates are next
made, which had better enter the mould somewhat low down, in
order to have some head of metal at that point. The object of
this is to prevent air bubbles in the casting, by means of the
weight of in-pouring metal, whatever air there may be in the

Steam Cylinder in Loam.

mould being thus forced upward, where it escapes at holes there
provided, termed risers. The mouth of the pouring gate should
of course be a little higher than the top of the cylinder. Venting
is rarely necessary in loam moulding, except for such pieces as the
long S cores. Piece R, Fig. 34, is left for the purpose of receiving

TOP   Of P/T



FJUQ. 34. SZecurv CyMruter:

Screw Propeller.                              25

the scum, in order to leave the casting sound. Wherever the
molten metal is to touch the iron plates the latter should be
washed with loam.

Foundry Pits.—It should be understood that the floor we
have spoken of in the last example is not strictly the foundry
floor, but that of a pit deep enough to hold the whole mould.
Important castings like the one last considered, and especially
upright ones, are always thus treated, and after the mould has
been finished the space left in the pit is filled in and rammed so
as to bed the mould tightly against the sides of the pit, and thus
resist the pressure of the metal on casting.

A Screw Propeller can be best moulded in loam, a pattern
being provided for the centre boss. Referring to Fig. 35, a board
A centred on the vertical spindle, and balanced by means of a
small weight, is revolved so as to travel along the incline B c,
which is only a template curved so as to have D as its centre, and
forming part of a screw of the same pitch as the propeller. It is
very clear then, that by backing up the surface B c with loam, we
shall obtain a screw surface the same as that of the propeller blade
required. The next thing is to mark out the shape of the blade,
shewn in dotted lines. On the blade thus marked out, dried and
blackwashed, we now lay strips of wood, as shewn at£, Fig. 36,
representing the thickness of the propeller blade, and the surface
is then covered with loam up to this thickness, smoothed off, and
again dried and blackwashed, Now completely coyer with loam,
and so form top mould, which in its turn is taken away and dried.
The thickness piece being removed, the blade is completely
moulded, and this may be repeated for the other blades. Setting
all the lower moulds then in position on the floor, the bottom half
of boss pattern is applied (Fig. 37), and, being filled round with
dry sand at E E, the top half is treated similarly. Lastly, the
mould is completed by the addition of a core for the central hole,
and of the top box, and the whole has the appearance of Fig. 38.

A large Fly-wheel may be moulded without the necessity
of making a pattern for the whole of it A coke bed is first
formed on the floor for the purpose of venting, and a centre is
sunk for the spindle A, Fig. 39. Then the core box in Fig. 40 is
taken, which is formed so that a certain number of cores made

Fly-wheel Moulding.                           27

from it may reach quite round the outer rim of the wheel, as suggested
by the dotted lines; the back and top boards B and c being loose, to
remove the core, which may be made in dry sand. After levelling the
floor by means of board D, Fig. 39, the cores from Fig. 40 are set up
at E, with the curved surface inward and gauged from the centre by
the striking board F, which has the same radius as the outside of the
fly-wheel rim. A small space is, however, left for the application of a
coating of loam which is struck out at top and side by the board F.

We next require the arm cores. The box for these is shewn at
Fig. 41, and supposing we have in our case six arms to the wheel
this box must be made a sector of one-sixth part of the circle.
The top, bottom, and sides are removable, so that when the
box has been filled with compact ' dry sand' they may be taken
away, together with the rim part and boss, leaving the arm, which,
being tapered, may be knocked out with a mallet at G and so
removed. Some moulders might prefer loam for these cores,
which would be baked in the usual way.

Putting the sector cores in place, as in Fig. 42, a pattern is
used for the boss, with a cope of green sand at H. There only
remains the completion of the cope for the rim. This may .be
done in dry sand, contained in boxes shown in plan at j, and by
means of a pattern K placed in the channel formed for the rim,
top box j being put on and rammed up there. This pattern K is
passed round until the whole of the top of the rim is formed, and
is finally withdrawn by removing one of the boxes. The mould is
now complete, and it is only necessary to form the gates, which
should be pretty central, while risers (about four) are put in the
boxes j to shew when the metal has filled the rim, which is known
by its lifting a metal ball placed upon them. Of course great care
must be taken in finishing the mould, so that no unsightly marks
be left on the casting at places where the cores join each other.

Marine Condensers, being usually large cubical castings, are
built up in loam in the manner described for other objects, by the
aid of what are known as skeleton patterns, projecting flanges
having patterns and core boxes. (See Appendix //,,/. 782.)

Fig. 43 shews one or two objects suitable for loam moulding,
A being a large Air Vessel for a pump, and B a Cone Pulley
for some machine tool

Fzirtker Loam Examples.


The internal cores may be noticed in these examples. They
are in both cases supported by being hung from the top-plate.
In the air-vessel A, the portion a is struck out first, by means of

boards d and e, with a thickness piece, black washed as usual ; then
follow with internal core f, using board b. Remove / when dry,
and strike out g by means of board c.                                       l

The core g is now removed, but returned in company with fy
first taking away the thickness a a, and the whole structure is then

A L \


Wheel Moulding.                              31

"bolted together; h is a ring to stiffen ^ and / is a pronged plate
to support the core/

The cone pulley B may be struck out in the floor, while the
internal core is made separately on a plate : patterns being used
for webs and boss. In fact, any casting of regular shape, either
circular or cubical, may be moulded in loam much more
•economically than by means of wooden patterns ; the symmetrical
parts being struck, and projecting pieces having core boxes.

A few other examples of moulds in green sand for different
objects, requiring no special description after what has been
previously said, are given in Fig. 44, where A is a stop-valve, the
larger core box for which has a loose pin to form the impression
•needed to support the smaller core : B a large marine- or stationary-
engine piston, the core being supported by and vented through the
pieces aa, which are filled in on finished casting by screwed plugs.
Boxes are needed for the cores cc, and c represents the mould for
a plummer block. (See Appendix //, p. 784.)

"Wheel Moulding.—Not many years ago all spur and bevel
wheels were m.oulded by providing a finished pattern for the
wheel required, but as machine moulding is not only simpler, but
•far more accurate, and as it does away with the necessity for
storing heavy patterns, which are sure to be out of truth by the
next time they ^are- required, toothed wheels are now extensively
moulded by machine.

Scott's wheel moulding machine may be understood by
reference to Fig. 46, and it may be premised that three operations
are necessary in the working of it.

A board B is set upon the central spindle A (see Fig. 45), for
the purpose of striking out the greatest diameter of wheel, on
which the teeth are to be formed, giving at the same time the
height of the top and the bottom of the rim. The spindle being
•removed, the machine is put in the central socket c, Fig. 46, and
the operations are now to be explained.

A pattern D, of two teeth, is accurately made in hardwood,
and being fastened to the upright arm E of the machine, this
.arm needs to be—(i), fixed to the requisite radius of wheel;
(2), raised or lowered; (3), passed round the rim of the wheel
by the rotation of the arm F.

Spur Wheel Moulding.


The first operation is done by means of the traversing screw G
which slides the whole of F and E by the nut H, the latter being
fixed to the centre piece. The lifting and lowering is done by
the hand wheel jr, which by worm and worm wheel turns shaft K
and chain wheel L, and the sliding arm E is raised or depressed
by the chain which is fastened to each end and tightly wound
round wheel L. The third operation, the rotation of the arm F, is
effected by turning the shaft N from the handle M, and, the motion
passing through the change wheels is transferred to the worm
wheel o, which is fixed rigidly on the central portion of the

machine. Varying change wheels can be inserted in much the
same way as in a screw-cutting lathe, and the revolutions or
fractional parts made by the handle can be seen by means of the
graduated disc P, so that any part of a circumference, such as the
pitch of the teeth, can be accurately traversed by the mechanism
last explained. The teeth are then formed in the sand after the

;'     '                                                                           D

34                        -Bevel Wheel Moulding.

proper radius of arm is -fixed,"by lowering the tooth pattern,
rilling up with sand, raising, rotating the amount of the pitch,
lowering again, and so on until the whole wheel is formed. The.
machine is now taken away by attaching the crane chain to the
eye-bolt at the top of central spindle (see also bottom off. 58).

Core boxes are needed for the wheel boss and segments
between the arms, and for-central hole. Wood strips are used as
gauges to fix the cores in position and to preserve the proper
thickness of the rim and arms; and a cope being placed over
all, gates are formed and the wheel cast.

Bevel wheels are moulded in a similar manner to the
above, the principal alterations being the strickle boards and tooth
patterns. Fig. 48 will make this clear. A board A strikes out
the back of the wheel in green sand, and parting sand is applied;
an impression of the wheel back is then taken in the top box,
and the latter removed to finish; board B next forms bottom face,
cores and boss pattern completing the remainder. The core-
boxes for the wheel arms are shewn in 72^, p. 6t, and similar
boxes will be found in Fig. 42.*

^ Chilled Castings.—Where a very hard and durable surface
is required to a casting, u chilling " is effected by making that
part of the mould, where the said face occurs, of iron. When the
molten metal meets the surface of cold iron, it cools rapidly and
forms crystals of white cast iron, hard yet brittle, where it meets
the iron mould, and, fpr a .depth of* ,an inch gr> rriore within the
casting, according to the mixture used, or' the weight of the chill
mould; the rest of the casting is still grey and soft.

It would seem that the graphite crystals do not, under such
circumstances, have time to form, and so the carbon becomes
! combined with the iron. Suitable objects for chill casting
are:—rolls for plate mills used in forging plates (see Fig. 275,
p. 280), and which require a great depth of chilling, as they
have to be trued up again on being worn down; tram-car
wheels, the rim only being chilled, this class of traffic not
being capable of supporting the expense of steel tyres; points
of plough-shares, which wear away at a very rapid rate in
the earth; bushes for ordinary cart axles; railway chairs; pro-

* For other moulding machines see Appendices II., p. 785, and V., p. 969.



jectiles for large guns; and, in fact, all classes of work required to
stand wear and tear, and  not   especially  needing   machining.


Pig. 49 will explain the methods of chilling:—A, a car \vheel
rim; B, a* plate roll," and c, a shell. Rolls and bushes are
moulded upright, and share points made in an entire mould

36                           Malleable Castings.

of cast iron. The iron mould must be painted with a thin coating
of very fine blackwash before casting, and some care must be taken
in the forming of the gates, as, if there should not be sufficient
pressure from the ' head' of metal used, the iron will recoil on
meeting the cold mould, and form a rough casting. Care must
also be used in the case of bushes, to remove the core chill before
the casting cools dovyn firmly upon it. The chills (the name
given to the iron moulds) are usually made of good cast iron,
though, in some rare instances, wrought iron has been used. Some
of the details of chill moulding vary according to different autho-
rities. Some founders purposely rust their chills on being first
made, to assist the blackening in resisting the action of the metal,
it being generally believed that the latter tends to fuse and injure
the chill. Other founders, notably in the case of projectiles,
neither rust them nor use blackening. The chills should be
warmed before casting, in order to expel moisture, and should
have a weight about six times the casting they are to chill, or the
chilling will be .too slow.

v Malleable Castings are obtained by taking the article,
after being cast and cleaned up (this last is very important), and
putting it, along with others, in an annealing furnace, in company
with some substance that will absorb the carbon from the cast
iron. Such substances ?are, oxide of iron in the form of scales
from the rolling-mill, or some other of the metallic oxides, placed
in the furnace in a state of powder. The intensity of the heat,
and the time the casting should remain in the furnace, both
depend on the size of the casting and the amount of malleability
required, the usual rule being to keep it at a white heat for about
a week, adding to this the time required to raise the temperature
and to cool down, i

Fig. 490 shows an annealing furnace with cast iron boxes A A
holding the castings, which are covered with a layer of sand. Of
course, it must be understood that it is only to a short depth
below the surface that the casting becomes converted into
wrought iron, though right through for small articles.
^ Softening.—If a casting is so hard that it cannot be machined,
it may be softened by heating and cooling out in common sand,
or any other bad conductor of heat

Brass Founding.


Brass Founding.—The moulds used in the casting of
brass require no new description, the only difference in this
class of work being the manner in which the metal is melted.

As the castings required are much smaller than those we have
recently been describing, the brass is made directly in crucibles
of some,impermeable material, black-lead being the best. The
melting furnaqe is-shewn in Fig. 50, and usually there are several
of these side by side and separately connected with the chimney.
The top of .the furnace is only a little above the floor level, and
in brass foundries it is customary to have the part of the shop
near the furnace entirely reserved for casting purposes; the
casting and moulding shops being entirely divided by a wall
in the best establishments. The principal difficulty in the making
of brass is that of the different fusing points of the two metals
used—Copper and Zinc. Thus, copper melts at 1996° F.,
while the melting point of zinc is as low as 773° F.

The copper is first melted, and the zinc is only introduced a
short time before casting, by means of tongs, pushing it down in
small pieces under the melted copper. It should flare up on
doing this, which is a sign that the heat is quick enough. If it
is left in too long, much of the zinc will be lost by evaporation.*

* See also Appendix I., p. 747.

Brass Furnace.

a f/l

In bringing this chapter on casting and moulding to a close,
a few practical points will be mentioned, although it should be
clearly understood that perfect practice can only be obtained by
actual work on such articles as have been mentioned in the text

The Size of Gates, and the number of them, can only be
determined by constantly watching the results obtained from
previous work.

Flat gates «••• should be avoided as much as possible, as
they tend to clog, though sometimes they are beneficial when
they are required to break off easily without damaging the
casting. Fig. 51 is intended to represent diagrammatically the
different gates and channels used to supply a mould; the pour-
ing gate; the skimming gate, for the purpose of retaining the
scum (and here some ingenuity is required to keep the latter
in the skimming gate by centrifugal action, the whirling being
produced by admitting the metal at a tangent); the sprues or
connexions from skimming gate to mould (they may be of any
number considered necessary); the feeding gate or gates, the
use of which is to fill up any part of the casting which is likely
to shrink; and the risers, which are to allow the whole of the
air in the mould to pass out and so prevent blow-holes, the

Forms of Gates:                                 39

soundness of the casting being also in the hands of the pourer,
as he may keep the riser covered for a shorter or longer time.

The size of gates is determined by the fact that the metal
must neither flow too slowly so as to choke, nor too quickly so as
to break the mould. All the while the pouring is going on, the
moulder agitates the metal in the gates by means of an iron rod,
which he moves up and down until the metal has cooled so far
as to prevent him doing so any longer; in this manner homo-
geneous casting is more likely to result.

' A great deal of art lies in the
ramming of the mould, but as a
rule, the deeper portions of it should
be rammed most, as there will be
a greater head of metal on them.
The floor of the shop should be well
vented by a bed of coke and the
insertion of pipes, to take away the          ~ .                 -

gases  from all work  done  on  the                ^   "''"......'"

floor, and coke dust should be put

fn the joints of loam building; some cores too should have a

large amount of coke in their centre.

The venting of the mould is also a matter which, requires
a great deal of practical experience to enable it to be done
with success. Large cores, enclosed on three sides by metal
(' pockets,' moulders call them), should be particularly well
pierced, and green sand moulds should be much better vented
than loam or dry sand work, on account of the steam rising from
the damp sand and the compactness of the latter.

Cores require good support by means of plates or wires,
especially such as those in Figs. 25, 34, and 43, and all cores
that are not held down by the shape of the mould, should be so
fastened, for they are so much lighter than the molten metal,
that they would float out of position if left to themselves.

Cores should be dried in the stove for about twelve hours, and
should only be placed in the mould a short .time before pouring,
to prevent the absorption of moisture by them from the mould

Patterns are lifted out of the sand by screwing rods or
handles mto them (see Fig. l$d,jt>. 67), and raising slowly, at the


Moulders' Tools.


same time rapping the pattern carefully to prevent the sand adher-
ing, and a few points should here be noticed as regards the finish-
ing of the mould. The moulder uses trowels, and c sleeker*,'
(which last are only trowels of special shapes) to smooth away any
broken portion, but, if the mould is made too smooth, there Is
great danger of blistering or scabbing, from the fact that the
mould, having lost to some extent its porosity, refuses to allow the
escape of gas, and it is generally understood by moulders that
the hand makes the best trowel, though certainly it is always
better to let a mould remain, if possible, just as the pattern left it.
In Fig. 52, a few moulders' tools are sketched.

The upper box is usually
termed the 'cope? which also
applies to the outer mould, and
the lower box is .sometimes
called the 'drag* The rope
should be well weighted to
ensure sound castings. \Vhen
two boxes are used, they fit to-
gether as shewn in Fig. 2, and
can be easily replaced, but
if the floor serves as bottom
box, exact correspon- Vr>i*r
dence is obtained by
driving stakes into the
ground through the
lugs of the top box.
-^                    Cbaplets have been

TfaolS.         mentioned, and are

to stay cores that cannot be

otherwise   supported;  their use,  however, is not advisable,  for
they tend  to produce weakness in the casting in  which
remain.    They should be tinned, or at least freed from rust, to
ensure uniting with the casting.

We have already described the charging of the cupola;  it
remains for us to explain the method of tapping it    The
at the cupola is provided with two iron rods; one he         to
a hole-in the clay stopping, which he does as soon as the

Mixtures.                                 41

has brought his ladle under the mouth of the cupola ; the other
rod has a flat end, with a lump of clay adhering. As soon as the
ladle is full, he applies the clay to the mouth in order to stop the
flow. To tap with safety, he should stand on one side of the
trough and use his tapping rod obliquely, while on stopping again,
he should cut off the stream from the top side.

Mixtures of Iron.—This is another art which nothing but
observation and practical experience can reduce to a nicety,
different mixtures being used for the same purpose by different
founders; indeed, the success of certain firms depends in a great
measure on the mixtures used. Still, a good idea can be given
of what is required for each purpose.

The varieties of pig iron having been already stated, we will
now consider each separately.

No. i is the weakest but most fluid of all the pigs, and may
be used by itself for ornamental castings on account of the ease
with which it fills the corners of the mould, but it is usual to mix
it with ' scrap J to increase its hardness, ultimate strength, and
closeness of grain.

No. 2 is finer in grain and stronger than No. i, and is used
wherever some strength is required with great fluidity.

No. 3 combines the greatest degree of strength consistent
with fluidity, and is therefore most extensively used, and in great
favour with founders.

No. 4 is the strongest pig for foundry use (the remaining
numbers, 5 and upward, being only required for conversion into
wrought iron), and is therefore used for heavy castings requiring
strength, such as girders, columns, bed plates, &c.

For strong castings two-thirds of No. 4 may also be used with
one-third of No. i.

Scrap is the name given to the broken up parts of old
castings, which of course may be divided into good and bad
scrap. Some founders place great reliance on it, using nothing
but scrap with an admixture of No. i, say two-thirds of scrap to
one-third of No. i, while others prefer using an iron like No. 3,
mixing with it only a little scrap to strengthen it, and so produce
a harder, close-grained casting. It is also a good plan to mix
iron from different blasts.

42                               Steel Casting.

While speaking of scrap, we cannot do better than endeavour
to understand the advantage or otherwise of remelting. We have
before said that remelting is a disadvantage. It is true that the
iron becomes purer as regards the elimination of graphite, ac-
quiring a 'whiter appearance, with, at the same time, increased
strength and closeness of grain; but, on account of other im-
purities, it is no longer as tough as before, and its ultimate
extension is therefore decreased. Unwin says: ' Remelting im-
proves the strength, but if repeated too long the tensile and
transverse strengths suffer, though the crushing strength and
hardness increase.' (See Appendix //.,/. 779.)

For chilled castings, a strong iron, as Nos. 3 and 4, is needed,
because the chilling weakens the metal.

Malleable castings require a pure mottled iron, or at least one
having very little grey mixed with it; for if the particles of graphite
present in coarse grey iron are taken away in the furnace, honey-
combing will result.

Girders and columns must have a strong and elastic mixture;
cylinders should be treated for hardness as well as strength, and
therefore require as much white iron as convenient; pulleys need
a soft mixture, such as a large proportion of No. i with a little of
No. 3 for strength.

Steel Casting requires no explanation. Its conversion from
iron will be treated in a subsequent chapter. The only difference
in the foundry is, that in order to prevent 'honeycombing,7 which
has been a great trouble ever since steel castings were first used,
great care has to be exercised, and even then many castings are
wasted, while brittleness is only prevented by slow annealing for
over a week or a fortnight of time. (See Appendix /., /. 747.)

Sir Joseph Whitworth introduced a method of compressing the
steel while in the molten ingot by powerful hydraulic pressure,
in order to prevent this troublesome honeycombing, the only
objection to his process being considerably increased cost. (See
also bottom of p. 82, and Appendix ///.,/, 790.)


'' *f* • *»t fv*n]tltut 4»i| pattern making are so iwvil;it*H
v :K fir fatfi m tinker e%jirc ially requiring tfi know at l«*a 4
n*i»$fiti <«J fiitniiiiifi^t that if would be quite imjiosMblr I1*
flu i* !»H tmf i Itajifff» indrj't ndent ni <*a< h oilier, nn*t it**
4.rrl i/'f i^ill^iii iiitikiii^ ha** brnt MI intuit rtitiird «>» Hi
*! tf#rtr m Itfflr It?ff IH say as ftgauls llit: /<>/w* wlifi h
u »,h»rtiJ«$ tak**; but mh;iitn*fr has hi*en omitted W4* %%tll
M 4* ft iw !«* '4i|i|ily,

ii*i   iit,iti<»|*any  an    the   iwo    woods    iiisr»l
fl   *fe  kept   tfi huge  stores   until thf+|   iW
.itjil it ilnw deil air used tut' the larger «ml
•.» lint *ir«  not  hard inou^h for the lii'lfi11!
i ir r f *»if|iing as other woods.
rinti -.hfinks very httle itt flrytng, and »T* f MH
ii'i,    (lit aiifiuni of that fact, and lhis <%i%r

wi-%'1 r, pr«*f lint**"* it'* IPC for huge j«ittt%rns
.»tii^i 1*1 fitaiiof^ifty ; II %*ir|i% a little 111*111%
In nwifk

;IH* iiilw r mi-inrl"*

14 ftiuhi'f to f ntradri

ill fii|"it% as ttii'* will show u»« sfiiiir *4
U,r tr j< ttt*. ftff               iifi » failrffi in 4 jfiliii tilar \*ay.

I Jv* |ti^,«!fa *^wtM«l***ii  I'- IPPIW Ui»*             To jirrtriif Ilir

^           «.f ll<f nrr, ;*< At l}u  SJi ft i* *»»wn tijt into

tf#ff^ j»|            $w             mlr flic fwiiii                in  I*iff,  54,

ftmr.ri llit!                           itir

44                            Warping of Wood.

wood and, as a consequence, the centre plank is narrower at the
edges while the side planks become drawn in, in the manner
shewn at A. If the tree be cut into quarters, each quarter will
contract, according to the above law, in the manner sketched at
Fig. 55, the sector piece at B becoming narrower, the square at
c becoming rhomboidal, and the circle at D elliptical Even after

	"•• «,
 ~ t


	i i

// H(
	S » 1

	I i
	' ^
 5  'x

timber has been thoroughly dried, it will always warp If a

shaving be taken off it, as may be seen at Fig. 56, for a

surface is exposed, which, drying, must contract the

has been planed.    It is, therefore, suggested by some authorities

that, in first beginning a pattern, the wood should be

out and.cut to the size required, so as to have some little

to set before being used.    (See Appendix //., /. 787.)

Tools and appliances.—It would be unnecessary in a
like the present to occupy the student with elaborate
of the hand tools used by the pattern maker.    The following
trated list will serve to explain them :



f    *


i    *

.-  *.  t,

,  1!I       f   li

I  S   *

*».       **   I

i -•

;, .4.

46                          Wood-tiirning L athes.

The flexible plane is very useful for truing large concave or
convex surfaces, a symmetrical curvature being given to the sole by
.simultaneous adjustment at front and back.

Of machine tools, two or three lathes are required ; the first,
with a long bed and two headstocks, Fig. 59 ; the second, a large
face lathe for turning wheel rims, Fig. 60; and the third, a light
face lathe for small articles, Fig. 61. Their speeds must be con-
siderably above those used for iron turning. A tool rest is used
to them all, and the work is done entirely by hand, the tools
necessary having such edges as are sketched at Fig. 62.

The mandrel of the lathe is provided with a chuck, which has
a different form for each lathe; jthus, the face lathe has a screw on
the mandrel to receive the flange chuck (see A, Fig. 60), and the
face plate on which the pattern is turned is supplied by a disc of
wood bolted to the chuck. The lathe with the two headstocks is
provided with a chuck of the form shewn at B, Fig. 59, which is
well pressed into the end of the pattern, and so compels the
latter to turn with the mandrel; and in the small lathe, the
pattern is screwed on the mandrel at c.

Other machine tools required are :—a band saw, circular saw,
and, if possible, a wood-planing, machine.                                  i

Arrangement of Mould-—-It is the duty of the pattern
maker to so arrange the moulding of his pattern as to cheapen
it (the moulding) to the utmost, and give the least trouble in
the foundry.    Thus, if the mould is to be in green sand, as
little dried core work is to be \ used as possible, and very ofteta
a great deal may be done by? the introduction of loose pieces,
i which are left in the sand after the body of the pattern has been
; withdrawn, and are then removed in another direction (see A,
, Fig. 30).    He should put as litjtle of the pattern in the top box
as is practicable, for it is evident that this part of the mould
must receive the worst treatment, being lifted off and turned over,
perhaps more than once.   Added to this, the fact that the cope
has to be taken away so as to leave the pattern behind, means the
; using of a good deal of care, despite which much broken sand
\ may still appear; while the half pattern in the bottom box may be
; *| '                    ' lifted in full daylight, and no accident need happen to it.   We

*•'**                       may here mention generally the method of withdrawing trouble-




OH   >".'/


	I '
	H - M    .
 r-f-   p
 Laf    '      ^
 ^S   r**!     *
 jfl/ «      '          ^   ;

o w

i {*   . fi<

	/> ^ '    J f 4
f            •                          ^.".j           -  v    •    .,,,:^-.
	..... •" '   ' ........                     '                    1 r*«M


H •



some parts of a pattern by lifting portions of sand on
' drawbacks} An example often quoted, and a very goo
purposes of explanation, is that of a lathe bed.

Fig. 62a shews a section of the pattern lying in th
The bed is reversed, so that the planed surfaces shall be


scum; and these surfaces are in the pattern made loose piei
The upper portion of the pattern can be easily removed, fc
problem is to withdraw the pieces a a. This can only be ac
by building the sand at b b on plates cc^ provided with hand
which they and the sand upon them may be removed, afl

M/l                                                      ,{ M p                   |fi^            JMV * i

If                     pi              *ii|i til         iw| «i f <*t

!| »»|f

$.» *rfii*i«t*  p                         *j**f*m|4 i           P«{                   rf'i f »/*

j|H»f   Jtlff                  |J|!   If till   III, |«                                   #   ijj^fH     «-f

ff |»«»*|             ptf               '! w* n  11)*

44*1 JM*           *{ Htl lit* |<l fji«flfi.t         JM                               H|        ^Wi*

i«H |M i«f«4*t^m «*jtji'fjiti*f *iip |« |iftt«»iJF flit                 ^ w i Jt^

IM   f| »**t   it*     JRO   *4|l«jMl«i  M4l|i                l»ijrjt}            J4t!*f*   I'   f,t   ,';4nsf

--jt ii** r%**M* *    |ifi KH!». til  |fi|m»iftjiit, ,{|in                  ^ *'h t      1 ^|(

1*1 i|i«ijj|**%.    »tio MJ'ttj 40} ftflt* *|.*^|n I*i         l*«| ** **** -i« |^    ri

jrM*» |^  J *f ri*t%%«*j|i* III! 4A1!I{   (f|lt*i$f, jWHll|,«*4fl *Jt^ **) £ .»»"|v,*'   \\

4V »r,|»n
|f*   if^ljiiltf*     f*l  |4J|l*ll?t|   *i»M»»j   {|f   *»t             I f*ffl          */^i'f VAjx    *

tti |f4if4*fn^ * *»*|if »yi? ',>*v **»*^|J<{ »*!*'» *f4fiii4 *'!<»,! 'i»»*** t »*;*
^«|l ^4i'ffi fti A f.Mi»tAt )tMf,»i}|ff. ijii* jiiir H|«fi*f| |*.»^i»r *:* t
|!Ǥ *|ttM |t,#|{f *ti*id ivii] M til *iir4 .itjj^ 'luipptifi

tin* M| x«wj ,tin» i? |iiii  ^ti.ii

M XutTUI JtUI   4Hil!|J|iiHlt!   |{U M}|i
in   X.»l|U'l   »fij!   |M   .»* i *  *»tj;   t|

»»*}!   *<*Jt    *l|     fJIIflllf   Jf||   JH   S|^|

M.«j«. n\  p,i i!]|iM,i|ii^ « < »|    » ^   ,jf    ,      r
»v »»i<ii«»'   " r  '*.*«*»» j*»    *i>4 rrt   ',*<   4 ,
I ««fl ui »»«»*[     *;|.jifM:   ,   4      r     »  ii

pi  llftfl l.tjlp ,it|$ III  U At'*'. ;/v   ., ;   ^       a



Building Piilley Pattern.                        51

reasons which have been previously mentioned, and so, patterns
of the larger kind at least, are made of layers well glued together.
Fig. 63 represents the making of the pattern for the pulley shewn
in Fig. 10. The rim is first built on the face plate of the lathe in
the following manner :—

Pieces of wood of the form shewn at A being sawn to shape,
are truly planed on one side at least, and also at the ends, by the
help of a shooting board B, which is used as a guide; they are
then fitted together to form the first layers of the rim, by glueing
to the face plate, taking care to put a strip of paper between, which
is always done when work is to be afterwards removed.

When dry, this layer is turned to a true plane, and another
superposed in a similar manner, but so as to 'break joint/ and so
on; the whole is lastly turned on the face E, and, being carefully
removed, is reversed, and again turned on the back side. So
much for the rim. The plate of the wheel is next formed, as at c,
by halving one plank over another at right angles, these being
grooved to receive the filling quarters a a ; the plate is next bored
at the centre to receive the boss, and turned on the outside (see D).
A rabbet having been formed in the rim to receive the plate, one
half of the pattern is complete.

Fig. 64 shews the halving for a pulley, of six arms, each batten
being cut to a fraction of its depth indicated by the figures; and
Fig. 65 shows the method used for five arms, the boss being re-
quired for fastening purposes. These are for the arms themselves.

The upright ribs, used to strengthen such arms (see Fig. 12),
would be halved on their narrow edges, and the boss filled in by
segments as at E, Fig. 65.

Pipes.—Patterns for small pipes are made out of the solid,
but those of a large size are built up as in Fig. 65a. Polygonal
half discs are placed at either end, and at intervals. Upon these
are carefully fastened with glue (and screws if necessary) the pieces
forming the rim. The two surfaces A and B being made true
planes, the half pipes are glued together with paper between," and,
having dogs driven in at the ends c, are centred in the lathe,
Fig. 59. If flanges are needed the ends would be turned as in
Fig. 66, and the flanges fitted in the manner shewn.

Bent pipes cannot be built in this way,  on account of the





i         i          1                cfr               i          J



Building Pipes.

impossibility of bending each lath of wood; they are therefore
made as in Fig. 67, which shews a plan and end view, In the
example shewn, the pipe must be worked out by gouge and spoke-


Turning Quarter Bends.

shave, and tried from time to time by applying the template ;D ;
in fact, working in wood much as the moulder works in loam at
Fig. 25 in the last chapter.

A still more handy way of making quarter bends is to turn a
built-up ring of semi-circular section on the face plate of the lathe,
Fig. 60, and afterwards to cut into four equal pieces, as shewn in
dotted lines, Fig. 6yA. On removing and placing back to back,


F,Wj.  67 cu.

we have clearly two complete patterns, and flanges may be added
as necessary.

If a single bent pipe — of whatever form of bend (so that it be
in one plane) — is required, it should be moulded as at Figs. 25,
26, or 26a; the pattern maker will supply the necessary tem-
plates. Curiously shaped pipes, of varying bore, may have a
core box only ; the outer pattern being built by the moulder, who
lays thickness pieces on the core, smoothing off with loam ;
after taking an impression, the thickness pieces are removed (see
description to Figs. 25 and 26).

Joints. — Many other methods of jointing, besides halving and
rabbeting, are, of course, used, such as dovetailing and tenoning,
but we must content ourselves with a general notion of the making
of a pattern.

Varnishing. — When finished and sand-papered, the pattern

Core Boxes and Prints.


is carefully varnished so as to preserve it  from   moisture,   and
present a smoother surface to the mould.

Core Boxes.—The simplest kinds are such as are shewn at
Figs. 8, 10, and 14, where half the core lies in
each box. Pegs to unite them are formed by
knocking rough rods of wood through the steel
plates, Fig. 68, and then driving these into holes
in the core boxes (see a, Fig. 69). The pegs need
not have more than a quarter or five-sixteenths
of-an-inch projection, as, if they are longer, they
may stick. The exact correspondence of peg
and socket is found by pressing some little
object, such as a pinhead, between the boxes,
and using these marks as centres. Pegs are
also required to unite the halves of patterns. Wooden pegs
are now greatly superseded by brass dowels (Fig. 6&a).

The hollows of cylindrical core boxes are gauged by the use
of a property of the semi-circle—viz.,
that the angle contained by it is always a
right angle. So that the box may be
gouged out as in Fig. 69, and tried from
time to time with a set square.

More complicated core boxes have
been already drawn at Figs. 30, 31, 33,
and 41. The last one may be noticed as
a case of a box that must be loose on
every side in order to effett the safe
removal of the core.

Core Prints.—At this point we may
consider shortly the different kinds of core

Simple cylindrical core prints are shewn at Figs. 10, 15, 18,
37 : they require a slight taper in direction of withdrawal. Some-
times it is necessary for economy to core the bolt holes in the
flange of a steam pipe, especially if the holes are to be square
A little consideration here of direction of withdrawal will
show that, if we used prints, they would need to be of a very
special kind, so they are usually dispensed with altogether,

. 68 a^.

5 6                                   Core Prints.

and a template given  to the moulder of the shape shewn   at
Fig. 690.

The cores, of a length equal to the thickness (full) of the
flange, are pushed down to their place, and held there by friction.

But a case similar to the above might occur, when, on account
of the weight of the core or the accuracy required, it might be
advisable to have prints, and as plain cylindrical ones would not

draw, as previously stated, we are obliged to have recourse to the
1 tail' prints in Fig. 6gb. Here A represents the pattern with its
prints, B is the core box for the hole, and c represents the finished
mould The portions bb are to be filled in by the moulder either by
hand, or in the case of a difficult shape, by cores made from boxes.


Core Prints.


Yet another form of print is required, where the core can be
supported at one end only.   That part of the core lying in the print



GJUarudi wJJtfv BMJLajvb& PjyinL.

matrix has then the whole weight of the core to support, and must
in consequence be large enough for balance.   Fig. 6 gc will explain


sg                          Worm Wheel Pattern.

what is meant: where A is the pattern, n the rore l»«ix, ami
c the mould, the object being a 'dummy' gland for a strain

Referring again to the Worm Wheel in Fig. i2,tlK-- inrtti^i
of making the pattern will be understood by the help of H&s. 7^
71, and 72. It may be built in the way shewn for the pulley in
Fig. 63, and, after being turned on the rim, blocks of hard wuocl
are fitted on each half of the pattern, and glued in tin; nunm-i-
suggested at D (Fig. 72).

The outside surface of these blocks is now turned *<» as to
give a solid rim of wood, from which the teeth are to be rut.    To
do this a stud A, Fig. 71 (on the table of a wood-working marhincj
is fixed at the angle of the worm thread, and the wheel pattern st»t
upon   it, so that it can be rotated carefully the amount  of the
pitch., by gearing, much on the same principle as in a moulding
machine.    A revolving cutter B, driven at from 2,000 to 3,000
revolutions per minute, is advanced to the pattern, and ruts
the space between the teeth; the diameter*/of this cutler
be the same as that of the worm.    When this operation Is oint
pleted, the wheel is removed and placed on stud c, Fig, j 2.    Tin*
wrought-iron worm   intended to work with  the  casting,  l*t*ifig
marked with red ochre, is now advanced, together with itn wcwicl
bearings, to gear with the pattern, and the worm is rotated ;
wherever a little mark is left by contact of the worm, the            is

gouged away until a perfectly correct fit is obtained.

Spur Wheels too small to make by machine moulding
have their teeth formed by the revolving cutter shewn at it,        of
course, in that case, the axes of wheel and cutter are at right

For machine-moulded wheels, either spur or bevel, the
is to be supplied with a block of pine with two teeth dovetailed in
harder wood, as in Fig. 720 (the machine is shewn at Fig. 46),

In both sketches the direction of withdrawal is shewn by the
arrows, and it will be seen that, although the bevel teeth
without difficulty, there would be some risk of the sand               to

the pattern in the case of the spur teeth, which are made
perpendicular and without taper.    To avoid such an                 a

finger bit A is provided, which, fitting in the hollow               the

two teeth, presses down the sand as the block is lifted


Bevel Pinions.


In Fig. 72/5 we have the core box required for the arms of a
machine-moulded spur wheel; its description will serve also for
bevel wheels. A represents the casting to be obtained, having
six arms, and the box at B is so designed as to core out a space of
one-sixth of that within the wheel rim. The box being in the

foundry is placed on a true table a, and after filling with loam, is
smoothed off with straight-edged batten at b. Six of these cores
being dried and blackwashed, a pattern for the boss of the wheel is
now necessary to complete the mould.

Small Bevel Pinions require the patience of the pattern
maker. Referring to Fig. 73, which is the section of a bevel
pinion, it will be seen that the teeth vary in size from A to B,
and must, therefore, be entirely gouged out by hand. The body
of the pattern is carefully turned as at c c, while blocks D, for
the teeth, are planted on in hard wood and again turned, as in
the last example. The section of the tooth now being set out by
compass or template at A and B, the teeth have to be cut out and
finished by hand. The teeth at B are made correctly lineable
with those at A by means of the wooden spindle K, carrying a
straight edge F so cut as to be always truly radial when moved
round the surface A B.


; ^

j: —

1 t

* 1


i tr f

*i  3

^   ? *? **

62                      Allowance for Contraction.

Contraction of Castings.—This is a subject involving
both thought and practice, and although a few general rules can
be given, success depends on very many points. It has been
previously mentioned that the moulder raps the rod that draws
the pattern from the sand. This rapping taking place in a

horizontal direction, it is evident that the sides of the mould
only are affected by it.

The pattern maker must not only take account of this, but
also of the particular moulder he has to deal with, for some
moulders lift a pattern with less rapping than others. In small
castings, up to about six inches across, the enlargement of the
mould by rapping will be about compensated by the shrinkage of
the casting; but in large moulds, the amount of .shrinkage will be
so much greater than the effect of rapping, that the latter may
be entirely overlooked, account being taken only of the in-

Plate Moulding.                              63

•crease in size of pattern necessary to compensate for contraction.
Patterns two to three inches across, or less, should be made
.about By smaller to allow for rapping only, and as this does not
take place in an upward or downward direction, there should
always be full allowance for contraction at these places.

The greatest shrinkage due to cooling will usually occur where
there is the greatest body of metal, and use must be made of
this knowledge by the pattern maker.

The linear contraction for different metals is as follows :—

Cast Iron      ......     |-"    per foot     =     -125"

Brass.........    T»/         „          =    '187*

Gun Metal    ......      -j"         „          =     -166"

Steel.........    fV"         „          =    -187*

Malleable Cast Iron         -f5/         »          =     '^f

Aluminium    ...         ...    \Y         „          =     '265"

Spur wheels about 2 ft. 6 in. diameter, contract B\" per foot,
and such wheels vary their contraction, increasing to •£$" per
foot for a wheel 10 ft. diameter (Box).

Three-foot rules, longer than the ordinary rules by the above
fractions, are called Contraction Rules,' and are used by the
pattern-makers, but with much care and judgment.

When wooden patterns are made, from which are to be
moulded metal ones, a double contraction should be allowed
•on account of the two mouldings necessary to produce the re-
quired casting in the first case, and the consequent double

Metal Patterns are required for light work or when a great
length of service is required. Such patterns are usually the same
as the wooden ones from which they are made ; but there are
•other examples of moulding with iron or brass patterns, as in
Plate Moulding. This is handy for such small articles as
•occur in a brass foundry; Fig. 730 will shew the method. A
wrought-iron plate a is provided with half patterns on either
side, made in brass and carefully finished. Prints are run for
connecting each pattern, so making channels for the flow of the
metal. The plate also has corners b />, so that when put between
the boxes c c> and rammed up with sand, exact correspondence of
the boxes is obtained. Except for blackening and fixing of


pouring gate, &c., the mould is now complete, and will, no doubt,.
be admitted as economically made. Of course this method will
serve only when a large number of castings are required of the
same kind.


Stopping-off is a process which often serves to utilise a
pattern temporarily for a slightly different purpose to that first
intended, and so to effect economy, A simple example will

In Fig. 73^, a pipe pattern with flanges is shewn; we will
suppose a shorter bend is required.

All that is necessary is to %fix a flange on at A, and provide a.
stopping-off piece B of the same size as the flange, having a print
attached for the core.

c represents a plan of the mould, with the stopping-off piece
in position; the portion D being filled up by the moulder. The
rest of the moulding will be easily understood.

Chain Barrel in Loam.                            6 5

In the propeller which we moulded in our last chapter
(Fig. 35) a screw template was placed outside the mould. There
are?cases, however, when a screw is to be moulded in loam, but






where the  course mentioned cannot,  for   certain  reasons,  be
followed.    Such is shewn at Fig. 73^.

A Chain Barrel for a crane is formed with a helical groove
to receive the chain. A is the casting required; B shews the
striking out of the loam, and c the finished mould. The only
portion requiring explanation is the screw </, made of hardwood.
It is fixed to the bottom plate, and has the same pitch as the
chain groove, though, of course, a more abrupt rise (for this
reasc>n made as large as convenient). The striking board runs on
the screw, being supported by a roller, and balanced by weight, as
in the case of the propeller.

Quouin, B/zrs&l i


The rest of the mould is self-explanatory with what has gone
before, and is entirely formed by loam boards.

Rapping plates have become a necessity in order to prevent
injury to the pattern by the moulder.
They are shewn at Fig. 734 being let into
the pattern, and are screwed to receive a
lifting rod as there shown.
^ Crystallization of cast iron. —
During the cooling of a casting the
crystals arrange themselves in lines per-
pendicular to the surface, but the interior
portion, being cooled more slowly, pre-
serves its granular nature. Fig. 74 will
shew the appearance of a bar of cast
iron when broken longitudinally (the
student should clearly understand that
the markings are exaggerated).

If the corners of the casting are made
quite sharp the crystals will be abruptly
turned at these places, and, meeting each
other also abruptly for some distance
below the surface, namely, as far down
as they are formed, will create a line of

fracture or portion weaker than the rest. Whether these corners
be external or internal, matters not; -the same thing happens.
Fig. 75 shows other examples having 're-entrant angles/ as they
are called, A being a circular boss cast on a plate, and B a cylinder
with flanges. It will be clear that breakage would always occur
more easily at these sharp angles.

When the Menai Bridge was built, the hydraulic press made •
for the purpose of lifting the * tubes; had a flat bottom with
pretty sharp corners, as will be understood from Fig. 76, which is
a sketch of the press first used. Stephenson took the precaution
of building up at each 10 inches of lift, and, had it not been for
this, great damage might have occurred, for the bottom of *the
press suddenly gave way, and the tube fell through a space of ten
inches. Fig. 77 represents the press since adopted, the crystals being
allowed to take a gradual turn, so as to leave no line of fracture.

. 73d.


Re-entrant Angles.



It is a general law that there should be no abrupt corners in a
casting,  either external or internal, principally  for the  reasons

. 74.

- £/V TRA NT    A NGL.ES.

1111' =
	1 =
	i'llll Iff
 i ,    ,

i    2
	t zz.

already given, and also to permit of an easy flow of the metal, and
prevent the breakage of corners of sand.

Unequal Shrinkage.                          69

^ "Warping and Shrinkage of Castings.—The general
effects produced by unequal shrinkage during cooling should
be well understood in designing a casting. These may be pretty
well arrived at by the consideration that, other things being equal,
those places will cool last where the largest amount of metal is
aggregated. Our first rule, therefore, is to endeavour to keep the
casting uniform in thickness. For unequal cooling is sure to
produce internal strains, and that portion cooling jftrst will set, and
be compressed by the contraction of the fart that is still coding*
Besides, if a thin part join a thick part very abruptly, the cooling
may produce such strains as to break the thin piece away
altogether. We ought therefore to make the juncture of unequal
thicknesses as gradual as possible.

Take a Plate, Fig. 78, lying on a surface of sand. The top
part cools first, on account of being open to the air, while the
under surface is still in contact with the hot sand, and the effect
of cooling is to make the plate convex on the upper surface, by
the after contraction of the lower surface.

In a Hollow Cylinder, Fig. 79, the heat cannot pass-
through the core so quickly as it can from the outside, so the
latter cools first, and the cylinder is made barrel-shaped by the
contraction of the interior. We must also note that the outer
layer will be in compression (see Fig. 83), which is a cross section.

A Solid Ball will be found porous on the inside, if broken,
because the shell sets fir^t, and the internal metal, being thus held
fast, is bound to leave vacuities on shrinking.

A Girder of the form sketched in Fig. 80 will curve longitu-
dinally in cooling, for here the most metal is collected in the
larger flange, and the casting is therefore pulled together on that
side, after the top web has cooled.

A Pulley with a thin rim, as in Fig. 81, will cool last at the
centre boss, and so produce a compressive strain in the rirn ; if
therefore a piece were broken out at A it could not be returned.

Shrinkage occurs while the metal is cooling from a red heat
downward, and the moulder can do a great deal to prevent it
occurring unequally by uncovering at the red-hot stage those
portions of the casting which are likely to retain heat longest, and
by keeping others covered, for equal cooling means equal shrinking.







Hozv Avoided.                               71

Hollow cylinders of all kinds are better moulded by inserting
an iron tube in the core, through which cold water is allowed to
circulate, and this can be so regulated as to produce a tensile
strain on the outside metal if needed, or, what is better, no strain
at all.

The pulley previously mentioned can be improved by curving
the arms, as in Fig. 82, thus giving them sufficient elasticity to take
the strain off the rim ; and such an example as the girder must be
left to the moulder's ingenuity, the thicker portions being first
uncovered, so as to cool quickly.

Of course it must be understood that a casting is never broken
directly by compressive stresses, but only by tensile stresses
induced by them in some other portion of the casting. Thus the
pulley will break at the arms, in which a tension is induced,
rather than at the compressed rim, although the latter may be
thin. This observation is true for all cases, (See$. 1012.) '

(For Plaster Patterns see Appendix EL, p. 781.)



IT will be well, so as to avoid repetition in succeeding chapters,
to digress somewhat, in order to consider the" properties, and to
some extent the metallurgy of the materials used in mechanical
engineering, omitting only the consideration of their strength,
which will be treated of in the second part of this work.

These materials may be classified as follows:- —

1.  Cast Iron.

2.  Wrought Iron.

3.  Cast Steel.

4.  Forged Steel,

5.  Copper.

6.  Zinc.

7.  Tin.

8.   Gun Metal.

9.  Brass.

TO. Phosphor Bronze.
IT. Muntz Metal.

12.  Manganese Bronze.

13.  White Metal.

14.  Wood.*

But we must first become acquainted with such chemical
elements as are necessary to understand the processes we intend
to consider. Such are: Carbon (C), Silicon (Si), Iron (Fe),
Sulphur (S), Phosphorus (P), Manganese (Mn), and Oxygen (O).

Carbon is an allotropic element, that is, it exists under
different forms, which are: Charcoal, blacklead, and diamond.
The first is pure carbon, and so is coke, or nearly so. The
second is not lead^ and is also called graphite and plumbago \ and
the third is the crystalline form. If carbon is allowed to unite
with oxygen it forms carbon dioxide (COa), a gas. Carbonic oxide,
or Carbon monoxide (CO), is another gaseous combination of
carbon and oxygen.

A chemical combination is the union of elements in such a way
that they can only be separated by chemical action, while a
mechanical mixture requires only mechanical means (very often
filtration) to break it up.

* For further materials see Appendix II,, pp. 794-801.

Cast Iron.


Silicon exists in combination with oxygen as silicon dioxide
or silica (SiO2), and is so found in the crystals of sea-sand; glass
is a mixture of several silicates. It is of value in cast iron.

Iron is found in combination with oxygen, the ore being
termed a ferric oxide (Fe2 Os), but it may be rendered quite pure
by chemical and mechanical means.

Sulphur is well known in the form of brimstone, and is an
impurity in iron, producing red-shortness.

Phosphorus is also an impurity, producing cold-shortness,
while Manganese is of value only when mixed with iron and
other metals in certain definite proportions.

(x.) Cast Iron.—There are seven varieties of iron ore, con-
taining from fifty to seventy per cent, of iron in their composition.
The blast-furnace (Fig. 84) is used for smelting the ore, which is

Jl  Xees-SjLdbe/


done at a very high temperature, with coke as fuel, and lime or
clay as a flux.*    The molten iron is run into pigs, while the slag

* Lime is the usual flux, but clay is sometimes required, as in the case of
hematite ore, and then is applied in the form of clay ironstone.


Blast-furnace Action.



formed by the combination of the flux with the impurities of the
ore, is separately withdrawn.    (See Appendix //., p. 788-9.)

The action in the blast-furnace is this :—Air being introduced
by the blast to give us oxygen, and coke to provide carbon; then,
if the coke be heated to redness, carbon dioxide is formed,

Air.     Coke.

20      +      C        :


As this gas ascends it takes up carbon from the coke, which it
passes on its way, thus :

Carbon                  Carbon

diox.      Coke.     monox,

C02 + C =  2CO

And we now have carbon monoxide.

Ascending further, this last-mentioned gas meets the iron ore,
which is now at a great heat. The oxygen in the ore has then
the choice of remaining where it is (Fe208) or of combining with
the C 0 \ preferring the C 0, it forms with it carbon dioxide once


Carbon                             Carbon

monox.          Ore.              diox.          Iron.



And the iron is now left, but in a viscous condition. As it takes
up carbon, however, it becomes more fluid, descends to the
bottom of the furnace, and may be then run out

Other substances have also been absorbed, which may be seen
on reference to the table at the commencement of Chapter I.,
shewing the general composition of the different pigs — grey,
mottled, and white.

Sulphur produces red-shortness in cast iron, that is, makes it
brittle when red hot, and Silicon and Phosphorus cold-shortness,
or brittleness when cold.

Carbon increases fluidity at the expense of strength, and
Manganese seems to have the effect during the smelting of in-
ducing the combination of the carbon with the iron, thus tending
to prevent the formation of graphite.

(2.) Wrought Iron. — The white pigs are broken up and
subjected to the processes of refining 2ci\& puddling. As, however,


Puddling.                                 75

these are chemically the same, and the preliminary refining is
very often dispensed with, we will give our attention simply to
the preparation of wrought iron by puddling.

The object of puddling is to eliminate the graphite entirely,
and the combined carbon so far as to leave only about '25 per
cent., which actually increases the strength of the iron.

In the rolling mill, where the metal is rolled into plates or
bars, scales of oxide of iron (Fe2 Ga) are formed by the contact
of the hot iron with the air. These scales, being broken off,
are collected for the puddling furnaces, their use being that
of absorbing the carbon from the iron, exactly in the way already
described for malleable cast iron.

Being intimately mixed with the brokenf white pig in the
puddling furnace, Fig. 85, and subjected to a powerful flame, the
O from the oxide unites with the C of the iron, and passes
away as C 02 gas. Any silicon that is present in the iron
unites at the same time with some 0 and forms Si O2, so that
the iron is left comparatively pure. During the process, the iron
is in the form of a spongy mass, and absolute contact of it with
the scales of oxide, now liquid, is ensured by the introduction of
a long rake through a small opening in the side door, for the
purpose of stirring the whole well together.* As the puddling
nears completion, the metal is kneaded by the rake or paddle

* To avoid hard labour and increase the output, there are many mechanical
furnaces now in operation, notably Danks' Rotary Furnace and the Pernot
Revolving Hearth.

76                               Puddled Bar.

into balls or blooms, and these are then removed and compressed
under a steam hammer by rapid blows, so as to squeeze out the
slag. The blooms are next rolled out and further squeezed by
being passed through the rolls of a rolling mill, giving us iron
called Puddled Bar.

These bars are now broken up and re-worked by hammering
and rolling, more or less, depending on the degree of purity and
strength which is required, and we thus have the varieties of
wrought iron known as—common, best, double best, and treble best,
which are used for various ordinary forgings, while Low Moor iron
is required for the fire-boxes of steam boilers and for more
difficult forgings.

The purification of the iron obtained in a puddled bar is
shewn by the following table, which may be compared with the
table showing the composition of white pig (p. 2) :—

Table showing chemical composition  of Puddled Bar, in

Iron         .........        ......        99*31

Combined Carbon         ...        ...        ...           -3

Silicon     ...        ...        ...        ...        ...           '12

Sulphur   ...        ...        ...        ...        ...           -13

Phosphorus        ...        ...        ...        ...           '14


Wrought Iron during its conversion from the pig, has lost
the capability of being cast Into moulds, but has acquired a new
nature, becoming viscous or sticky, and, as a result, maybe worked
by the smith, when white or red hot, its formation into different
shapes being assisted by the property of 'welding, which as cast
iron it did not possess. Repeated rolling gives a fibrous quality,
making the iron both stronger and more homogeneous or uniform
in texture, and these fibres may be seen on breaking a bar of
rolled iron, which then has the appearance shewn at A, Fig. 86,
while cast iron or even puddled bar gives a granular fracture (B).

Rolling or hammering iron when cold or nearly so gives it a
crystalline structure near the surface, so that T iron is not so
strong as bar iron, and plates still weaker. Re-heating and slow
cooling tends to remove this source of danger. (Seep. 1013.)

Composition of Steel,


Generally, then, wrought iron is tough, and more capable of
resisting vibration than cast iron, its fibrous character giving it
also a distinct advantage in the direction of the fibre, which
property may be made use of by judicious crossing in the opera-
tions of piling and re-heating the iron after puddling.

The best forgings are usually made by the piling of wrought
iron scrap.

(3 and 4.) Steel is intermediate in composition between
-wrought and cast iron, thus:

Cast iron may have  2-2   to 5    percent, of carbon.
Steel (for casting)        -3    „  17     „         ,,

Steel (for forging)       -25 ,,1-5     „          ,,

Wrought iron             'o    „    -25   „          ,,

It will be clear, however, that the exact limits between which
we   may   call   the   substance   £ steel,3
without intruding on either wrought or

- cast iron, are very difficult to define, so
that we may have steel which is almost

. as brittle as cast iron, or we may have
it on the other hand nearly as soft as
wrought iron.

Although steel has an intermediate

• composition, it has not, as we might
. expect, an intermediate tenacity or use,

' but is stronger even than wrought iron
. and consequently more useful. It never has quite the toughness
. of the best wrought iron, though approaching it closely with
...mechanical treatment; it is always more homogeneous,

It will also be readily           that, as           is intermediate In

Cementation Process.



composition, it may be made either from wrought or cast iron.
We shall first consider the former method.

Cementation.—In this, the oldest process, bars of wrought
iron are placed in fire-clay boxes, Pig. 88, with charcoal dust
around and between them, and a layer of fire-clay over all
(being the cement giving the name to the process). They are then
subjected to a bright-red heat, for a time varying with the amount
of carbon required to be introduced, and which may be as much
as a fortnight for the more highly carbonised steels. The charcoal
then becomes combined with the iron, and the steel so pro-
duced is called blister steel, from the fact that the bars are covered
with blisters. These bars are next broken up, piled, and heated
in a furnace almost exactly like the one in Fig. 85, hammered by
rapid blows from a tilt-hammer, Fig. 89, and shear steel of a

fibrous quality is thus produced.* Double shear steel is made by
breaking in two and re-hammering Crucible cast steel is obtained
by melting fragments of blister steel in covered crucibles made of
a mixture of fire-clay and plumbago, and placed in sets of six or
twelve in furnaces having a similar section to the one shewn in

* The steam hammer is used in later-built works.      For drawing,  see
Chapter IV.

Bessemer Process,


Fig. 50. Several of these crucibles are poured simultaneously to
form the ingot, many well-drilled workmen co-operating to do it
carefully. This variety of steel is much more homogeneous and
has a greater tenacity than shear steel, having a fine granular
structure. Brittleness is corrected, and the property of weld-
ability restored by the introduction of manganese in the form of
carbonate of manganese.

The Bessemer Process is used for the purpose of obtain-
ing steel from cast iron. Fig. 90 is intended to shew the neces-
sary plant employed. The converter A is filled with molten cast
iron, and air is blown through the metal by means of the tuyeres
at the bottom. The O of the air combines with the C of the
iron and passes away as C O2 gas, leaving the mass as pure iron,
the silicon forming a slag (SiO2) on the surface, which is


Open-hearth Process,

separately removed The temperature must be exceedingly high
in order to preserve the iron in its fluid state after the expulsion
of the carbon; the entire absence of the latter is discovered by
the application of the spectroscope, this being the most practical use
of that most wonderful Instrument. The next operation is the
adding of so much carbon as is needed to produce the steel re-
quired, and this is done by putting into the converter a measured
amount of very pure cast iron called Spiegtleism* and mixing it
well with the metal by re-applying the blast for a short time. The
now converted steel is transferred to the ladle B, which is swung
round by the crane c, and the metal poured into the ingot through
the hole D on releasing the plug at the bottom of the ladle.

The ingots may be afterwards piled and rolled as previously
described, to produce a fibrous steel, and if used for forging and
welding purposes should not have too much carbon in their com-
position; or, if required for steel castings, may be re-melted in
suitable quantities, much as in the way already mentioned for
cast iron.

The Siemens-Martin, or open-hearth process, is carried on
in a special kind of furnace, called a regenerative furnace, invented
by Sir W. Siemens. Fig. 91 is a drawing which will shew all the
necessary parts. A is the hearth, sloped in the side elevation, so
that the metal may run out when tapped at T. A current of air is
allowed to pass under the hearth at c, to prevent the melting of
the iire-clay. The combustion of a mixture of common coal gas
and air is the source of heat, the arrows showing the passage to
the interior of the gas through the valve o, while the air enters
through the valve A. In the figure the mixture is seen entering
the right side of the furnace. Being ignited at j by means of a
red-hot bar, gradually and carefully at first, the flame is directed
by the roof on to the metal, and the heat passes away by the left
side of the furnace, returning through the valves and past the
damper D to the chimney. Were it not, however, for Siemens'
beautiful regenerative principle a great deal of heat would be
wasted. The regenerators are shewn at nurr; they are hajrd
fire-clay or silica bricks piled as a grating. The rejected heat
from the hearth is intercepted by those marked n R, so that
* See Appendix II, j>. 790.

Regenerative Furnace,


although the mixture enters at 700° F, the products of comhustion
pass to the chimney at 200° or 300° F. In a short time the bricks
become white hot, and the valves A G are then reversed as is
shown at v1 and v2, the former being the position for the action
already mentioned, and the latter allowing the gas and air to

? ^g^e/te^xg^

. 9 £

enter first the left Me of the furnace and leave on the opposite
side. Doing so, it is evident that the heat which was absorbed in
the last operation by the regenerators R R is now taken up again
bj the entering gases, and the bricks rr, in their turn receive the

rejected heat

By this means a large amount of heat is made useful which

82                    Whitworth Compressed Steel.

would otherwise be wasted, the valves being reversed reguh
whenever the bricks acquire too much heat.

The furnace is first charged with pig iron, and when thii
melted, heated  wrought  iron and  steel  scraps  are  added
degrees (these three in nearly equal proportions).    When all
thoroughly mixed  a  little  piece  of cast  iron, in the form
spiegeleisen, is added, together with a very little manganese.

Experience, the principal guide for this mixture, is again call
into play immediately before completion of the operation, t
foreman trying small samples taken from the furnace and cool
in water, by breaking them and examining the fracture. If sat
factory, the steel is now poured into ladles by tapping at r.

As soon as the metal ceases to flow easily it is known tr,
there is only slag left.    The ladle is then removed, and the si
,*                         allowed to run to the ground or into moulds.

1                              The Landore-Siemens process, also the patent of Sir 1

t|                         Siemens, differs in the fact that iron ore is used direct   On beii

|ij                         first reduced and the slag got rid of, it forms spongy balls ,

tf                         malleable iron,  which are then dissolved in molten  pig  iro

,|1                         spiegeleisen being added as before.    It often receives the nan

l\,                         of the * pig and ore ' process.

n                              In the Siemens process the ore and flux are mixed dire*

with the pig; more slag is therefore produced.

Steel Castings made by any of the above methods mu,
be annealed slowly in a closed  furnace for a week or mon
to prevent cold-shortness.     Honeycombing,  or the presence  <
vacuous spaces in the metal, is the principal trouble, and is parti
, |                          prevented by the addition of silicon, as silico-fenromanganes*

I                          but; is only perfectly got rid of by the Whitworth process, whei

f                          the molten ingot is compressed by powerful hydraulic pressuij

until it is «quite set.    The great advantage of this compressioil
which amounts to from six to twenty tons per square inch, I
'shewn by the fact that the ingot is made to contract as mud
f as one-and-a-half inches per foot of length.    The mould consist
' of a steel cylinder, lined with refractory material, and so con
' strutted that when placed under an hydraulic press, the gasei
may escape through the sides of the mould.   (See App. //.,/, 790.1
1    We may always expect highly carbonised steel to be deficien!

Copper.                                       83

in toughness, and therefore inferior to wrought iron in that
respect. It may be improved by the annealing spoken of, but
steel that is required for boiler or bridge work must be capable
of resisting vibration, and so a milder quality is used, which,
though it may not be considerably stronger than iron, is more
homogeneous and has a finer grain.

The amount of carbon varies with the use to-which the steel
is to be put, and is shewn by the following table :—

Razor tern Sawfile Tool Spindle Chisel Sett Die
	Per ......
	iJj °l  Carbon .....
	Will spoil with over- heating. To be heated only cherry red. May weld with great care. Ditto. Tough ; will harden at low heat. Stands hammer ; welds easily. Stands pressure ; welds like iron.

		if/.   „   ......
 1*7     .,     ......
	i   7
		i'/.    »    ......
Cutting tools require most carbon.    (See Appendix //.,,/. 793.)

Tempering or the capability of receiving any degree of hard-
ness, is a property of steel, and was formerly applied as a test
to distinguish it from wrought iron; while case-hardening is a
method of partially converting wrought iron into steel, but both
these subjects will be reserved for our next chapter.

Test.—A rough test to distinguish between wrought iron
and steel is to put a drop of dilute nitric acid on the
polished metal, when a greenish-grey stain will indicate iron,
and a black spot will shew steel; the purer the black, the
more carbon may be suspected, so that we may even get a notion
of quality.

(5.) Copper Ore is various in character, but may have iron,
sulphur, antimony, or arsenic associated with it. The operations
are three in number:—(r) Roasting^ to get rid of arsenic and
sulphur, the iron forming an oxide. (2) Smelting, to dissolve
the iron oxide, and leave copper combined with sulphur. (3)
Roasting and Smelting, to remove the sulphur and obtain metallic
copper. The furnaces used throughout are of the same class as
the puddling furnace, Fig. 85, and called rewrberatory on account
of the arch beating back the flame. Other refining processes have

Gun Metal,

to be gone through before the metal is considered fit for the

The metal thus obtained is rolled into plates and hammered
to any shape. Besides its malleability it is exceedingly ductile,
being easily drawn into wires; it becomes brittle if hammered
cold, but its tenacity may be restored by annealing.

Copper is an expensive material, and is only used for pipes
that require bending cold, and for fire-boxes, where ductility as
well as power to conduct and resist heat are needed: it must be
remembered, however, that copper loses its strength somewhat
with increase of temperature.

It is also very useful for electrical purposes, being, next to
silver, the best metallic conductor. (See Appendix If., p* 793.)

(6 and 7.) Zinc and Tin are of little importance singly to the
mechanical engineer.

(8.) Gun Metal is an alloy of Copper and Tin, and is often
called bronze. The proportions are varied for different purposes.

Thus to make 100 parts :—


Soft gun metal requires 90
Hard gun metal ,, 82
Bell metal             „ 80


I o (General Ordnance purposes.)


Usually some zinc is added
malleable, as : —

Zinc     ,

to   make   the   metal   more




Gun metal produces fine castings, and being muck stronger
than cast iron, is almost the only other metal preferred
besides cast steel, for the castings required In modern
gunnery. It is often in its harder farm made into bearings
for shafts, Both strength and toughness are increased by rapid

(9,) Brass is roMe by alloying wpper with tifa   The pro-

Bronzes and Brasses.                           8 5

portions vary somewhat, depending on the colour and strength


Parts Copper %         Parts Zinc p/0
Fine yellow brass has       ...    66*6        ...        33*3

The proportion of copper may vary from 66 to 70 per cent, or
even higher. A little lead is sometimes added. Brass is principally
used on account of its fine colour, and because it is easily tooled.

(10.) Muntz Metal is a brass having the proportion of 60
per cent, of copper and 40 per cent, of zinc. It is largely used
for bolts in marine work that are liable to rust, and especially for
pins that have to turn in their sockets, on account of its great
strength, as well as the faculty of being forged, which it possesses,

(n.) Phosphor Bronze is, like gun metal, a mixture of
copper and tin, but with the addition of a small measured quantity
of phosphorus. (See Appendix /,/. 748.)

Its strength is then so much increased as to be equal to that
of wrought iron, and it has consequently been extensively used,
within recent years, where strength is required, coupled with
intricate form, such as must be cast rather than forged; as for
example, toothed wheels subjected to shock. Gun metal is
deteriorated by subsequent meltings, while phosphor bronze may
be re-melted without injury.

It has considerable ductility, and may be formed into wire, and
used for spiral springs subjected to steam or water.

(12.) Manganese Bronze, introduced later, is really a
yellow brass, to which about 7 Ibs. of cupro- or ferro-manganese
have been added per cwt. of the metal. The strength is thereby
still more increased ; and it is used now for a variety of purposes
where strength and ductility are required combined, such as
hydraulic pipes, which may be then drawn considerably thinner
than copper ones; and it is advantageous in many other cases,
as may be understood from the fact that it way be both cast into
moulds and forged under the hammer. It can also be used to
rust, so as to keep nuts and bolts free that would otherwise
seize. (Set App. //., p. Sot.)

< (15,) White Metal, otherwise white brass, and in America
Babbitt's Metal, or * Babbit/ is an alloy used for lining bearings.
Tin is the principal metal used, and is mixed with copper and

S6                                   Brazing.

in             proportions, the following percentages being-


Copper   .          . .        ...        ...        •••      °       3

Tin         ...        .........        -.    84     9°

Antimony          ...        ..         ...        •••      8        7


of white metal for bearings is that it can be

run         the                     the journal is in place, and so ensure a-

fit.     It causes considerably less   friction  than brass or

To sum up then, alloys of copper and tin are termed bronzes,

and                a little zinc added up to about i \ or 3 per cent.

of           and 2inc are called brasses, Muntz metal being

one of          ; and those having tin and antimony, with a little

are          metals.

^   Brazing.—Brass or gun metal may be united by this process,
is also termed hard soldering;  and the joint will then
be as            as the original casting.

or steel  may  be also connected  by brazing if  more

especially finished pieces of work.    The method is to

first                clean the work with acid, then take some brass

mix with powdered borax as a flux, the borax being

moistened with water.*   The filings are placed between •

the           to be braced so as to form a joint, as much surface

for the latter as possible, and the two are held together

in             tongs, having thick jaws to keep the heat.    The tongs

will         the          and grip the pieces until perfectly set,- and the

be finished off in the vice.

If the work cannot be easily gripped, another way is to insert
the           as before, and, binding with iron wire, place the' pieces '

in a          coke fire until the operation is complete.

Or, still another method is to use the blow-pipe.    Here a fine
of very hot         is directed on to the work by blowing with' •
this instrument through a lighted 'Bunsen.'          *                 •   •

*            ge&omfy^ Instead of brass Slings, 'spelter* is used, which' is a

about equal parts of copper and zinc, and specially'

ft* bntafug pturposes*                        .                              ...   :-   '

Wood.                                     87

(14.) WoQ,d is not used to so great an extent as formerly.
Roofs are made of wrought iron; and men-of-war of iron and
steel instead of oak: pillars of cast iron : while morticed wheel
teeth are almost out of fashion. Brake blocks, too, are made of
cast iron, to give a longer time of wear -3 and wooden buffer beams
for locomotives are now being discarded.

Little then need be said of wood. For pattern-making, as
already stated, jbtne, mahogany, cherry, sycamore, lime tree, and
walnut are the woods used. JLnglish oak is the best for beams,
but American oak is much cheaper, and the latter is used for the
framing of railway and traction waggons, and for locomotive
buffer beams. Ash is also much employed in waggon work,
especially for cart shafts. Mortice teeth are made of beedi or
hornbeam* Lignum vitas is of great service for bearings that are ,
immersed in water as, for example, mth the screw-propeller and
some turbines. (See Appendix 77~.,/. 787.)

Railway sleepers are rendered very durable by impregnation
with creosote or black oil, air being first sucked from the pores of
the wood. The creosote is then forced in at great pressure.

The following table gives the melting points in degrees
Fahrenheit of the principal metals mentioned in this chapter :—

Cast Iron   __.....    2ioo°P.

Wrought Iron____    3000° *

Steel   ...____.....    2700°


Zinc   ..............      773°F-

Tin.................      44V

Gun metal......—    1900".

Brass......   1700° to 1900°

v   Soft Soldering.—See Appendix K, ^.970,

* Castings of * -wrought iron' have been-made^' though   the process .is-
somewhat   intricate,   and has not ,been extensively , applied.    The method .
consists principally in lowering the high melting point of wrought iron by the
addition of aluminium.    Swedish wrought iron is used,' and from T
of its weight of aluminium is mixed with it,                               "


WROUGHT iron is formed into the required shape by drawing
down, and bending while hot; but if there should be insufficient
"' stuff,' or if it should be more difficult to entirely finish by
drawing down, recourse is had to welding.

The working of de-carbonised iron may be best treated undei
two heads, smithing and forging.    The first includes the making
| : i                     of  such   smaller   objects  as   can   be   conveniently   done    at

a smith's fire, while the second term may bd applied to the
shaping of all articles that require heating in a close furnace, and
finishing under a heavy steam hammer. In either case the result
is denominated a forging.

l| ;                   •                                  THE SMITHY.

«1 ,                            We will first consider shortly the plant and tools employed by

11 ',            •          the smith.

The Hearth,—This is represented in Fig. 92. A is a sectional
elevation, and B a front view. It is necessary to explain here that
the smith may arrange his coal on the hearth in two distinct ways,
the one being called an ' open' fire, and the other a * stock ' fire.

||                         The hearth shown in Fig. 92 is by Messrs. Handyside, and is ofi

iron throughout. It is only adapted for i open ' fire working,
being short in length from a to b. a is the tuyere or blast nozzle,
constantly surrounded by water contained in the tank ^ so as to
avoid burning at the outlet, or the accumulation of caked slag.
The work to be heated is placed in the hollow portion of the
hearth surrounded by coal, and as the coal burns away more is
supplied from the hillock £. It will then be seen that there is noi
special difficulty in arranging the coal for ' open * fire working.!

:( I                        ' Stock' working requires a certain amount of trouble in first pre-i

1                           paring the coal, which is usually done first thing every morning.!

After this first preparation it will, however, keep in working order!



9O                               Smith's Fire.

for the rest of the day, and has many advantages, as will be seen
Fig- 95 represents an ordinary smith's hearth, built up partly o:
brick and partly of iron, a is the blast nozzle, which need noi
now be surrounded by water, because the fire will never be nearei
to it than the position marked b, and so no caking can happen,
In building the ' stock' a loose brick is first taken out at c, and a
bar passed through and inserted in the tuyere. The coal is now
damped by sprinkling water upon it with a wisp of straw, and is
built up into the form shown, the ridge d being neatly flatted
down, by using the back of the shovel. Beginning at the tuyere
and advancing frontwards the * stock J is finished round the piece
of wood <?, which is called the t stock block.' We may now
remove both bar and block, and make the fire in the space e.
The iron to be heated is placed in this space and covered up with
loose coal, which is always brought from the front end <:, so
that the stock gradually burns away to the end b by the close of
the day. The advantages of ' stock' working are these: (i) we
need no water tuyere nor consequent attention to water supply;
(2) the bar to be heated is only acted on by the fire to the length
required (whereas ' open' working has a tendency to heat it to a
.greater extent); and the method is generally more economical.

The Blast.—Air is constantly supplied to the fire, when
working, by means either of bellows, fan, or blower, one of the

*|f!                           latter two being in use at an engineers' smithy, where all the'

fires are connected to one main blast pipe. Fig. 95A, Plate I.,
represents a fully equipped smithy, as designed by Messrs.
Handyside, and fitted with their hearths throughout The

iS                           main blast pipe is shewn by the dotted lines in plan.

Fig. 93 is a drawing of a Fan l^jr Sturtevant. There should
be a good large space left beyond the vanes, to allow the velocity
energy given to the air by them to be easily transformed into
pressure energy in the pipe, and so prevent waste by eddies.

Roots' Blower, as made by Messrs. Samuelson, of
Banbury, is shewn at Fig. 94. The air in this machine is

iff                          literally scraped out of the casing on the side A, by the revo-

/ ;                           lution of the two figures s s, in opposite directions, and is delivered,

at B, a fresh supply replacing the partial vacuum formed The
rollers, as the above figures are called, are compelled to work



(fj&fodi wxJttv

'5           20           25           30           35


flan, for Smithy





>i    t*






«*»>*.%* ./   /T4MMK t fWr***^


( n

V       'ft

f*        *,*x


•• A

Smiths Tools.


in the side elevation. The power absorbed in running this
machine is very slight, and the speed need not be more than 300
revolutions per minute. A fan, on the other hand, to be effective,
must be driven at a great velocity, say from 1000 to 2000
revolutions per minute; more shafting and pullies are required,
as shewn in Fig. 93, and the percentage of loss by friction is
consequently high. The blower is, however, very noisy.
# Tools.—Among these we must first mention the Anvil,
Fig. 96. It is made of wrought iron, and has a surface of steel
about a quarter of an inch thick welded on at A to form the top
face; B is the beak or horn; c and D are square holes to receive
* bottom' tools, and E E are used in punching. At Fig. 968 is
illustrated a French anvil. It is not provided with any holes,
the swage block (described later) serving instead.

Two kinds of Hammer are required: the hand hammer
weighing two-and-a-half to three pounds, for the smith; and the
sledge hammer, used by his helper, weighing from eight to four-
teen pounds, and even more. If the sledge is only worked by lifting
over the shoulder, a short handle' is used, say three feet long, but,
when swung, in making heavy forgings, a long shaft is required,
the right hand being drawn inward to the end as the hammer
approaches the work, thus giving the latter the full effect of the
stored energy.

Other tools, shewn in Fig. 97, are principally for the purpose
of finishing work for which the hammer alone would be in-
sufficient They often go in pairs, as top and bottom tools, the
smith holding the first by means of a hazel rod wrapped round it,
while the second is placed upright in one of the square holes in the
anvil. A A are chisels^ B B fullers, c is a flat-face orflatter', D a punch^
and E E are swages. The last term is applied to any specially-
shaped top and bottom tools designed for the purpose of finishing
work with greater ease and accuracy to a particular form, such as
round, hexagonal, &c» F is a set hammer, having either a square
or circular face; it is held steadily on the work while being struck,
so that in one sense it is not a hammer at all It is convenient
as a top tool to reduce work or ' set' it down, the anvil serving
as bottom tool G is a 'heading* tool, useful in making the
heads of bolts and pins. It is held by the hand at one end





•/< <//;>*/


4 JN


;«!»^  *

-*   ,

*\ f



* V       4           »

' Ml .

*     ,tt*     '

i       J


HAZ.CL   #00


Steam Hammer.


are cut off of proper length to form the rivet, and being heated, are
dropped into the hole at b. The hammer h is now worked from
the footboard/ the blow being delivered by pressing the foot down-
ward on the latter, while
the return of the hammer
is ensured by the elasti-
city of the sapling of ash
s, which is bent on each
down stroke of the foot-
board, and in becoming
straight again lifts the
hammer. The correct
form of the rivet head
is given by applying the
cupping tool <r, held in
the hand. When the
rivet is finished it may £/LCf. «V.y.
be released and thrown
out by striking the foot
sharply on the lever /,
which thus takes the
dotted position, and the
rivet can be then picked
up and cooled in water.
i Steam Hammer
for Smithy. — Lastly,
the smith requires for
his heavy forgings the
aid of a small steam
hammer. We say* small'
to distinguish from the
larger type in use by the
forgeman,but the smith's

hammer is anything but small The one illustrated in Fig.
Plate IL, is spoken of as a 10 cwt steam hammer, and this means
that the piston and piston rod A A, the tup B, and the pallett c,
together weigh 10 cwts. This, of course, does not take account
of the steam pressure, which at 40 lbs« per square inch con-


r f t
1 I f


> >

^  z   *»   * %

y  X   —    *  **

*t           n    f-

«*    *                   "Z

»"•     --ft,      -,.      •»

|f Cf

5  fe ^  *

?j_   ^   *«»,   ^

r  r  % ^

**  t,  ^

*   f    y  s

^ <*
** «. *
*i *>






Steam-hammer Valves.


form. In the drawing, steam is shewn entering by the mid port,
and passing down the lower cylinder port to raise the piston; at
the same time the exhaust steam from the top of the piston is
passing down the upper port and out at K. Now (supposing the
self-acting lever H is out of gear), the piston having reached its


***v»*            ir^i| 0jrt| f

tm ,^|| jp.. ||a4 S'ip^f;  <





102                                 Welding.

resorted to, the crystallised portion will be left weak and little
better than cast iron. This should be carefully noted in making
connecting rods of steam engines, or indeed any article the break-
ing of which might cause danger to life.

Welding.—Wrought iron cannot be cast,* but it can be
welded without difficulty; that is, it may be joined piece to piece
by heating and hammering, and work of great intricacy may thus
be formed. The welding temperature for wrought iron is reached
at about 2800° Fahrenheit, and the two pieces to be welded are
heated to this temperature, which is detected by the iron beginning
to throw out sparks. Two points have to be noticed. The iron
should be, if possible, drawn out so that a scarf may be made,
when welded; this is shewn at A, Fig. 103, and, as will be seen, a
greater surface for welding is thereby presented. But, if it be
drawn out too fine, it will burn away when put into the fire for
the welding heat, and to prevent this it should be left rather thick
at the ends, as at B ; the lump may be easily levelled afterwards.
The two pieces to be welded should both be at their proper heat
at the same time, which the smith ensures by changing their
positions in the fire, so as to advance the one or retard the other.
Withdrawing, he sprinkles them with sand, which forms a siliceous
film or flux, and prevents scale by oxidation. Putting them now
together, the smith gives one or two blows to fix them, and he and
the striker then finish by rapid alternating blows. If the flux be
carefully expelled and the joint well hammered while hot, the bar
will be nearly as strong there as at any other section. Borax is
used as a flux in steel welding. (See Appendices /. and Iff.,
pp. 748 and 917.)

The scarf weld is the one most commonly practised, but the
fork weld at c, Fig. 103, is often introduced for large work on
account of its greater security.

Having thus briefly mentioned the operations of heating and
welding, we shall now proceed to describe the forging of a few

The making of a Bolt with hexagonal head is shewn in
Fig. 104. A round bar A is taken, of suitable length; it is
heated at one end, and jumped or upset, namely, is lifted by

* See note at end of Chapter III.

IO4                       Forging Bolt and Nut.

the tongs %and struck t>n the anvil as at B. A heading tool is next
held over a hole in the anvil, and the piece B is reversed and
dropped through the tool. Being prevented, however, from
passing quite through, on account of the shoulder just formed, it
is now beaten by the hammer until the head c is formed. The
bolt is then taken out, and the portion c is roughly hammered
into the form of a collar at D. It will now have become cold, and
must be re-heated to finish the head, which is done in the hexa-
$i                        gonal swage E, side after side being presented to the tool by turning

jjl                        the bolt round, and hammering each time.    Finally, it is dropped

into the heading tool once more, as at F, and, after receiving one
or two finishing blows, a cupping tool / is applied to give the
spherical chamfer.

We may now make a Nut for the above bolt Of course, it is
almost unnecessary to state that bolts, nuts, and rivets are now
made entirely by automatic machinery, and these examples, there-

f jf                       fore,  are only intended as an   introduction   to   more  difficult

forging.    A nut can be forged in the form of a ring, and thus

j|li                       dispense with after-drilling.    This is the case we shall consider.

\\                     Fig.   105 illustrates the different operations.    Slightly scarf the

bar A, which is to be bent round to form the nut, and must, there-
fore, have the same width as the latter; for example, a three-
quarter inch nut would require a bar about three-quarters of an
inch by three-eighths of an inch in section. Next heat the end of
the bar and bend round the anvil as at B, nicking it through with
a blunt chisel (as shewn at a in sketch c). Now, put it back in the
fire to get a welding heat; take it out; and, breaking off sharply
at a, lift up the ferrule remaining, on a mandril D, and weld the
two scarfings together; then finish the hexagon in the swage E.
The nut is not yet complete, however. Re-heating, it is cupped
at top and bottom as at F, and the hole is finally made to
exact size by the finishing mandril £, which is driven through
the nut into the hole h in the bottom cupping tool. The nut
may now be removed and cooled.

Fig. 106 shews the making of a Holdfast for pipes, or pipe
hook. Two heats are necessary. In the first a bar is taken, as at
A, and is drawn to a s square' point on the further edge of the
anvil as at B, a turn of 90° backward and forward between each

io6                            Forging Eyes.

blow being given by the hand holding the tongs. A second hea
is now taken, and the length of point having been marked off (c
the remainder is set down at D, on the edge of the anvil Hen
again the bar is turned backward and forward to finish the edge*
in plan E, and the end is chipped off at e to proper length
Before the work is too cool the part e must be bent round the
beak of the anvil, as shewn at F, when the holdfast is complete.

A Single-Eye in the form of an eye bolt is shewn finished
at A, Fig. 107 (page 103). The hole is to be drilled out afterward
A short piece of round bar is first taken of the same diameter as the
collar, and after heating is fullered at B, and set down as at c. Or
the second heat the edges are hammered, and the corners chipped
off with chisel as at D, shewn in plan. One end of the eyebolt is
thus finished. Taking a third heat the line E E is marked off, and
the tail of the bolt swaged down at F. Finally, the shank is cut
off to the required length.

We will now describe the forging of a Double-Eye. A in
Fig. 108 gives the finished form, serving as the end of a tie or
connecting rod, to which it is welded when required. A square
bar is taken (exact length of no importance) rather thicker than
the part marked ay and is first heated, jumped, or upset as at B,
and then flattened out in swage block till it assumes the form
shewn at c. Being heated a second time it is drawn out as at D,
partly on anvil, and partly by returning it to the hole in the swage
block, when it is finished off at the ends by chipping off the
corners shewn at E. A third heat is required to bend the T thus
formed round the anvil beak to the fork shape F, and the fourth
and last heat will serve: first, to hammer out the octagonal por*
tion; and, second, to swage out the round part H.

A Pin with cotter is our next forging. After heading at the
first heat, like the bolt in Fig. 104, it is then of the form BJ
Fig. 109. On the second heat it is cut to the length required,
and the cotter hole marked off. The latter is c driftedJ through
by means of the tool c—first, with the work lying in a bottom
swage; and, second, to finish—by driving the tool through, over
a hole in the anvil, see D. In punching and drifting the tool must
be kept cool by taking it out of the work, and dipping in the;
water tank from time to time. A represents the finished pia The

jg./O9,        JRin &, Cotter*


Forging Spanners.

cotter E           little description.    It may be formed by bending

a thin strip of Iron as at F, welding the portion near the bend, and
chipping out the narrow shank.

The student will have already noticed that a good deal of
Judgment has to be exercised by the smith in deciding upon the
length and breadth of iron necessary to execute a certain piece of
work, and. although this can rarely be achieved with very great
nicely, yet practice enables him to guess it with sufficient accuracy.
As a role the cubic contents, or the weight of the stuff should he
about the same in the rough as in the finished piece, some
allowance being made for burning away In the fire, but it is best
to err by having rather too much tban too little, and in most
articles the extra stuff can very easily be cut off. Some, however,
require more exact measurement, as from the nature of their con-
struction the after cutting cannot well be resorted to. Wherever
parts have to be afterwards machined extra material should be
allowed, say from one-sixteenth to one-eighth of an inch, but the
careful smith will always leave as little as possible, and if he is
directed to finish *&lack3 he should make the work as exact to
dimension as his tools will allow.

Except in the case of the nut at Fig. 105, none of the work
already described has called for the operation of welding. We
shall now, however, pass on to some examples requiring the aid
of this Important process,

A common S or Double-sEnded Spanner is the article
we shall first consider. A, Pig. in, shews the finished forging.
A bar Is taken of the same length as the arm, leaving a little
extra material for welding. It is heated and first bent to the S
form (B) on the anvil beak, straightening by flat hammering on
the face of the anvil; -and is next drawn out at the ends as at c.
Now, two pieces of rather thicker bar being procured to form
the jaws, these are heated and bent round the beak, and the
comers chipped off and rounded as at r>. Heating again, these

are finished on the bottom tool and scarfed down as shewn
at ?, We are now ready to complete the 'spanner by welding the
Jaws to the arm, at the searings already made (see &), and finish

be given between the flat face H, and the anvil..
In Fig. na, A represents a Shackle for use with chain or

Forging a Shackle.


rope. Some little care should be exercised in gauging the length.
For an 'inch' shackle, made out of round bar one inch in
diameter, a length of fourteen inches would be required. This




bar is set down to the form at B, by using the set hammer and
bottom swage &. Two heats, one for each end, are required for
this purpose. Another heat for each end enables us to make a



Forging a Shackle.

scarf of the form shown ate, by drawing down at the point and

The eyes are next formed  by taking a welding ^ heat,

round a mandril rather smaller than the finished size of

and velding with hammer as at D.    A flat face, E, is used

to smoothen, and a finishing mandril is driven through the hole

as at F.    These operations being performed on each eye, we have

fee shackle admnced to the stage G.    Daly one more heat is

necessary to bend the rod to the* proper form round the

beak3 and the finishing stroke is given on a block (H) which

as a template to define the distance between the eyes.

Forging Hooks.                              ill

An JCyebolt of large dimensions is treated in Fig. 113. A
is the finished condition. It is such an eye as would be required
for the attachment of a rope or chain, being made of round
section to prevent cutting or chafing. Here we may begin by
taking a round bar of the same section as the part A, and, wrap-
ping it round, scarf and weld it to the form of the eye as at B,
at the same time scarfing down the joint again. This done, a
second bar c of thicker section is cut to form the shank, and,
after scarfing, is welded to B, giving the appearance D. Lastly,
the collar is put on by taking a piece of square bar of small
section, which may be wrapped round the shank at welding heat
and scarfed at E. The bolt is then finished off by fullering the
part b, and swaging <?, a rough file being used with advantage

Another and probably quicker way of making the eyebolt is
to take a bar of the same diameter as the collar, and work out of
the solid by swaging down the shank, fullering and flatting out
the eye portion, the hole being punched and rounded off.

As an interesting example of punching and swelling out we
may take Fig. 113^. Here we have a portion of a Harrow-
frame, and it is desired to form the socket for a common square
tyne. The bar at A is first upset, punched, and drifted to the
form at B. It should be noticed that at first only a narrow, long
section of drift is used, to avoid breaking the bar. The narrow
hole is swelled into a round one by a suitable tool on the next
heat (shewn at c), and the final step is the further swelling by
square drift, as at D, carefully finishing with a flat-face.

Hooks may have the eye formed in the manner described
for the shackle of Fig. 112, or the large end may be 'jumped/
and worked from the solid by means of a flat-face tool, either in
the case of hook or shackle, and the hole left to be punched or
drilled cold. The -solid method needs no special description.
Assuming a case similar to the one previously described for the
shackle, the bar being first round and of the diameter of the
thickest part required, the eye end of the bar is drawn to the
proper diameter for that place, while the opposite end is drawn
down nearly to a point This is shewn by sketch A, Fig. 114.
The eye is next turned and welded, and the hole finished with


Forging a Box-key.                           113

mandril either now or afterwards (B). Heating the rest of the
bar the hook is bent to the correct form round the anvil beak c,
being constantly checked by rule and sheet iron template; and the
proper section given at the same time (shewn at r> D) by means
of set hammer or flat-face. Both these last-named operations
must go together, for the form of the hook will be more or less
spoilt by flattening to the section at D D, and this must be again
restored by bending.

Bolts in machinery are sometimes placed in very extra-
ordinary positions, so that the spanner in Fig. in may have to
be discarded, and the Box-key (represented in Fig. 115) used
in its place. It has a socket at A to fit the nut, and a shank at
&, on which a wrench (sketched at c) is placed when required.
The key is forged by making the A and B portions separately, and
afterwards welding them together. Thus, part A is made by
bending a strip of iron, which has been previously Scarfed at the
ends, into the form of the hollow cylinder D. This is done on
the anvil beak, and a second heat is necessary to weld it. The
piece B is next formed from a round bar of sufficient section to
give the square when flattened. It is shouldered on a swage as
at E, sufficiently small to fit into the ring D. And now the small
end of E and the cylinder D are both heated to welding tempera-
• ture; then, being" put together as at F, are riveted by striking the
mandril G, and by hammering round as at H. The fourth heat is
required to work out the square j with flat-face and anvil, and on
the fifth and last heat a mandril, which may be hexagonal or
square, as desired, is driven into the cylindrical portion K, and
the outside hammered until the Tequisite shape is given to the
hole. Removing the mandril the key is considered as finished.

Tongs, having to be used almost continually, are soon burnt
away by the fire, and the smith must be able to forge them as
needed. We will therefore describe the forging of the round-nosed
tongs sketched at B, Fig. 98. The ' bits? that grip the work are
made-first. For them a piece of square bar is to be set down on
the edge of the anvil until it receives the form A, Fig. 116; the
successive operations for this are shewn at i, 2, 3. The two bits
should not be made right and left-handed, but exactly alike, for
in turning one round axially it will be found to accommodate


Solid Forging v.  Welding.                    \ 15

itself quite correctly to the other. One heat should be given for
each of these settings down, and during the third the hole (B) is
punched. Next, the handles are to be welded to the bits, and for
this purpose round rods of sufficient length are scarfed, heated to
welding, and united in the usual manner c, being finished care-
fully in round swages, D. The nose bits are yet flat; they are
therefore rounded by means of atop fuller and bottom swage, as
at E, and, finally, the two half-tongs are riveted together tightly
as at F with a hot rivet, the handles being worked backward
and forward while the rivet is cooling, and also during the
after quenching in water, This method ensures a well-riveted
but workable joint

The student will notice that in the processes of forging two
principal methods are followed, which in many articles merge
considerably the one into the other. These are the forging of the
object (i) entirely from the solid, by drawing down or cutting out;
and (2) the joining of the parts of the forging by welding. The
former is a process of cutting out or carving, the latter of building
up. Figs. 104, 106, 107, and 108 are examples of the first method,
which is the one practised unless the method of welding should be
cheaper, and, as we shall see, is always used if possible in large
objects that have to sustain important loads. Figs, in, ir 3, r 15,
and 116 are cases where the second method is more useful, for in
Fig. r 16 a round bar is attached to work that is easiest forged
from a square bar, and the end pieces in Fig. 112 are manifestly
easier made separately and welded, than they would be by forging
completely from the solid,

Further examples of welding are shewn in Fig. n6a. In each
case A is the work prepared by scarfing or otherwise, and B the
built up article. The Eye may be said to be merely an example
of ornamental welding, for it would be difficult to find a use for it
in practice, The Stud is more commonly met with.; It is prepared
as shewn, by scoring the surfaces to be welded with a chisel; less
pressure will then "be required, the form of the stud will not be so
much distorted at the shoulder, and tjie two pieces are much more
likely to enter into each other.

The next three forgings to be described will be worked in the
* solid' manner, and they will conclude our description of those



A          2





Jgtgr. //6V J^^zxa

Steam-Hammer Forging.


methods used by the smith.    They will also introduce the use of
tlie steam hammer, as applied ia the smith's shop,


Fig.   117'is   a   Single-webbed  Engine   Crank  shewn
finished at A.    A. slab of iron, is required of the same thickness

11S                            Forging Cranks.

and          as the            boss.    Heating to a good white heat, it is

put           the hammer, where the ferrule B stamps out the shape of

the             It is next drawn out by suitable tools, called sets, at

top and          (see c and B) until it is of correct length to form the

which is first set down to the proper thickness, and

by means of a ferrule, as before.    The forging is

now of the form E, and all that is necessary is to finish by cutting

corners round the bosses, which will require another

heat—the third ; the first having been used for the large boss and

the setting down, and the second for the small boss,

A Bell Crank Lever, whether large or small, can be made

in a similar manner to the foregoing.    A, Fig. 118, is the finished

lever,   A bar Is taken, as before, of the thickness and width of

the boss.    It is first bent to a right angle—if a small lever this

be done on the anvil beak, but, if large, blocks would be put

the steam hammer, with the hot bar between, as at B.   That

the boss is next formed by ferrule, as at c.   Another heat

will now be found necessary for each arm, in order to set down, as

at p D, to proper section, and the ends are finally cut to curve by

of a chisel or cutter (see Pig. 1.19*2:).

Figs 119 and uga represent the forging of a Small Crank
Shaft, say two inches in diameter, such as can be worked by the
with the aid of a small steam hammer.   A is the finished
and has two crank arms forged upon it at right angles to
other, in the manner of locomotive  axles  for   'inside'
cylinders.   We mast, to begin with, have a slab of iron of square
section, sufficiently krge to form the crank web when  drawn
down,   This is seen at B.   It should also be long enough to
complete the whole shaft when drawn down and swaged in the
manner to be described    The bar B is first to be formed into the
shown at c, by heating to a good white heat and setting
down under the hammer, as at D.   TMs will leave the slab of the
section as the crank web, and, if carefully set down to the
Indicated, the webs will now be in correct position, namely,
at ijgbt angles to each other.   Of course some care must be
the right angle being tested, with a square, and the part
m £ la jttzticalar should be made of such a length that when
* .to:therotind -section it will measure-the correct distance




%/r f *>


f    *       f





122                                Stamping.

between the crank arms. Probably this piece had better be
swaged next (it may require another heat), the forging being
turned round, backward and forward, to produce a good result
(see E). The distance between the cranks should be now finished
very exactly, by knifing or other means. The ends remain. Here
it is necessary to first cut out the superfluous material by marking
off at F3 punching the hole o, and, while the crank is still hot,
cutting out the rectangle with a knife or cutter (see H). After-
wards the shaft is rounded by swaging (j). When this has been
done for both ends, and the shaft carefully measured, as well as
tested for axial straightness, straightening if necessary, the work
may be considered complete. In this form of crank (double-
webbed) the pair of webs are always forged solid in the manner
described, and the piece between taken out either by slotting or
turning in the lathe,

A.t this point we may as well consider one other form of crank,
which has many advantages. In Fig. 120, A is the shaft alluded
to, and is there shewn finished by turning in the lathe. It is con-
siderably stronger than the one previously described, on account
of the fact that the fibres follow the bend of the crank webs
(represented in dotted lines), while in the shaft of Fig, 119 these
fibres are cut through when the mid pieces are slotted out, which
must of course weaken the webs considerably. The only
objection to the form here shewn is that a great xridth is required
for the crank itself, and, as this cannot always be spared, the
crank has only been applied on portable or traction engines up to
the present Properly we might have described this in the space
devoted to the forge, for a larger hammer is required than
commonly occurs in the smithy. A bar of the best Yorkshire
iron, of sufficient diameter to torn down to finished size, is heated
and placed between the blocks B B, and these are made to
approach each other by blows from the hammer, at first gently,
and afterwards more strongly. Lastly, the shaft must be tested
for stxaightaess.

Stamping*—Where several articles are required exactly alike
in form and dimension they can often be forged more cheaply
by the use of stamping tools, The crank last described might
almost b€ termed an example of this kind of work, and the lever

I             H


q. f21.

I                           124.                                Case-

'                        in Fig.   117 could be stamped by means of the tool shewn in

Pig.   121, the hot iron being placed in the hollow H, and the
:                        hammer brought down upon it,    The ragged portions are after-

wards chipped off the forging. Usually these stamping tools are
made of massive cast iron, but if they are to be used extensively
cast steel will be found necessary. Other examples of work
suitable for stamping are shewn in Fig. 1210, where A is a
spanner, B a double eye, c the centre portion of a screwing stock,
D the handle portion of a lever, and E the boss part of the same
lever. (See Appendix //-,/. 802.)

,                              Before leaving the smithy two processes should be explained,

because they are as a rule performed by the smith.    These are
•;                        the methods of hardening wrought iron and steel.   Cast iron, as

j,ji      "                        we have seen in Chapter 1, can be easily hardened at the surface

ij     •"                        by chilling, this taking place -while the casting is in course of

IVA                               formation.    Wrought iron and steel are hardened after the article

^                                is completed.

V Case-hardening.—This is the name given to the process
by which wrought iron objects are hardened to a depth of from
one-eighth to three-sixteenths of an inch below the surface. After
forging the work is machined and polished, and is then made to
absorb carbon by being placed in air-tight boxes or £a$es in con-
tact with some substance rich in carbon, being strongly heated
while in that condition. The method is much the same as that
pursued in the cementation process (Fig. 88), and it will therefore
be seen that the iron at the surface is converted into a film or
^                        case of steel, the only difference from the cementation process

being that the heat is merely kept on long enough to case tlie Iron
with steel and not to steel It qtilte through. WMlethe iron is left,
then, hard at the surface the inside remains tough, and is as
•capable as ever of enduring "vibration. The boxes may be either
made of sheet iron, or may be fireclay retorts similar to those in
use at gas works, and provided with a lid to keep them air-tight.
They may be heated as In Kg. 88, and the substance put In
contact with the iron is not wood charcoal, as in cementation, but
animal charcoal in the form of bones; for it is found, why It is
not quite clear, that if nitrogen be present the carbon will unite
more rapidly with the iron, Other substances may he used, such

Tempering.                                 125

as prussiate of potash, leather or hoof scraps, but the process is
chemically the same. After packing, which must be carefully
done, to prevent the articles bending while hot, the heat is raised
during two hours, the whole kept at a regular temperature for

about nineteen hours, and then allowed another two hours to
cool, Hernoving the articles they are quenched in water,
straightened, and re-polished. (See Appendices /., //, and F.9
/A 74», ^03, andwz.) ^

Steel of a mild quality may be hardened at the surface by the
absorption of more carbon.

Such small articles as have to withstand considerable wear are
case-hardened, <?,/., radius links for reversing gear.

Tempering is a method of giving to a piece of steel any

required degree of hardness.     Properly there are two distinct

meant when we speak of(tempering' a steel tool-    The

126                          Tempering Colours.

first of these is that of hardening. Here the steel is heated as
equally as possible to a c cherry red/ and not more; and on with-
drawing from the fire it is plunged vertically into a vessel of cold
water. The quickness of cooling has a great effect on the hard-
ness, and this may be accelerated by moving the article about in
the water. Cracking or warping will also be prevented by
judicious motion.

The steel is now so hard that it will scratch glass. It must
next be tempered or let down to the required degree of hardness.
If the tool be again heated to cherry red, and allowed to cool
slowly it will by that means have become annealed, and will be
at its softest; but if it only be heated to one of the temperatures
in the following table (Fig. 117^, Plate III.), and then cooled
rapidly, it will take a particular degree of hardness corresponding
to that temperature, and to be obtained at no other. When
letting-down, the softest tool will be that which is cooled at the
highest temperature, and the hardest that coaled at the lowest

The exact temperature which the tool has assumed is ascer-
tained by the colour which appears on the brightened surface, due
to a film of oxide of iron formed by contact with the air. There
is some difference of opinion as to the requisite hardness for
certain purposes, and slightly different colours are required for
different steels, but Plate III. is suitable for average tool steel.

Tempering a Chisel or Drill.—To make the matter
clearer we will take the case of a chisel for chipping metal. It is
forged out of a steel bar of the section shewn at A, Tig. 122, and
is drawn out (at as low a heat as possible, to prevent burning) to
a flat point as at B. This point is now to be hardened and
tempered, while the rest of the chisel is to remain in its natural
condition. Whenever the tempering is accomplished by quench-
ing in water, the preliminary process of hardening must always be
performed, otherwise the tempering would have no effect. In the
case of the chisel, or any too! having a point requiring a particular
temper, tbe two processes are performed at one heat, but it must
be quite clear that hardening is not therefore dispensed with.
Heating the whole chisel to a cherry red, the part a b only is
quenched in water, and so becomes very hard* Now rub Jthe
point of the chisel with a stone to brighten it a little, and, as the


f * * i * * I


* v i -;

'   f * t   «

126                          Tempering Colours.

\                        first of these is that of hardening.    Here the steel is heated as

equally as possible to a ccherry red/ and not more; and on with-
drawing from the fire it is plunged vertically into a vessel of cold
fj                              -water.    The quickness of cooling has a great effect on the hard-

I',                              ness, and this may be accelerated by moving the article about in

the water.    Cracking  or   -warping  will   also  be   prevented by
judicious motion.

The steel is now so hard that it will scratch glass. It must
next be tempered or let down to the required degree of hardness.
If the tool be again heated to cheny red, and allowed to cool
slowly it will by that means have become annealed, and will be
at its softest; but if it only be heated to one of the temperatures
,                        in the following table (Fig. 117^, Plate III.), and then cooled

rapidly, it will take a particular degree of hardness corresponding
j                        to  that  temperature, and to be obtained  at  no  other.    When

'•                        let ting-down, the softest tool will be that which is cooled at the

highest temperature, and the hardest that cooled at the  lowest

The exact temperature which the tool has assumed is ascer-
tained by the colour which appears on the brightened surface, due
to a film of oxide of iron formed by contact with the air. There
is some difference of opinion as to the requisite hardness for
certain purposes, and slightly different colours are required for
, |                        different steels, bat Plate III. is suitable for average tool steel.

I                   *         Tempering a  Chisel or  Drill.—To make the matter

\                        clearer we will take the case of a chisel for chipping metal.    It is

forged out of a steel bar of the section shewn at A, Fig, 122, and
is drawn out (at as low a heat as passible, to prevent burning) to
a flat point as at B,   This point is naw to be hardened and
•i  ' ^                        tempered, while the rest of the chisel is to remain in its natural

condition. Whenever the tempering is accomplished by quench-
ing in water, the preliminary process of hardening must always be
performed, otherwise the tempering would have no effect. In the
case of the chisel, or any tool having a point requiring a particular
temper, tlje two processes are performed at one heat, but it must
be quite clear that hardening is not therefore dispensed with,
Keating the wtok chisel to a duny ret, the part ab only is
quenched in water, and so becomes very hard, Now rub ihe
point of the chisel with a stone to brighten it a little, aad,--asr the




 > »
	:   1650°
	For Preliminary hardening only

H (L, O
	Too soft for anything

	Springs Screw drivers Circular saws for Metal Cold chisels for W.I Firmer chisels

H— (
	"   •       •  ••"•      "             ........ """•   •• " Cold chisels for CL Axes and adzes Cold chisels for Steel


	Flat Drills for Brass Twist Drills Plane Irons

	Gouges Reamers Punches and Dies

	Cha-sers Taps Screw-cutting Dies

	Boring cutters

<^ X
	Milling cutters Drills Iron-planing tools

	Steel- planing tools Hammer faces Light turning tools Scrapers for Brass

Fig. J.lf7b.—Tem,pe'rin,ff Table.

Face p. 126.


Examples of Tempering.


heat from the body b c travels down towards a, the colours will
appear, the point becoming gradually hotter, yellow first, then
through brown to blue. But we require for our chisel the tem-
perature of 550°, which is indicated by a dark purple; as soon,
then, as this tint is seen, the chisel is entirely plunged into water,
and the point is thus made of the correct degree of hardness. A
drill point may be tempered in a similar manner, using, however,,
the darker straw yellow for colour.

^ Tempering a Screw-tap.—Sometimes, when an even
temper is required over a considerable surface, the result may be
better obtained by putting the article in contact with a body oi
hot metal. Such a case is that of a screw-tap. The tool, being
finished and polished, is next to be so tempered as to make the

screw threads hard, while the square shank remains soft. An
iron tube being procured, A, Fig. 123, of such a diameter as to
just fit freely over the tap, the latter is first lifted by the shank,
within red-hot tongs, B. In the meantime the tube has been
heated to a dull red, and as soon as the first signs of straw colour
are seen on the tap shank, due to the heat from the tongs, the
tap is placed within the tube as shown. The .shank will have had,
so to speak, a start of the rest of the tap, and by tlie time c has
assumed a dark strawy colour, r> will have arrived at dark blue, a

128                Further Methods of Tempering.

higher position In the colour  scale.    At this point the tool is
quenched in water.

Two other methods of ascertaining the desired temperature
are in use besides the colour test.   These are the flashing tem-
peratures of certain oils, and the fusing points of certain alloys.
The first is practised by coating the part of the tool with oil, and
holding it over the fire until it blazes off, then quenching in water,
In the second, the alloys are usually of lead and tin, and vary
from equal parts of each metal to complete disappearance of tin
and consequently total lead.   A bead of the alloy placed on the
tool, may be watched until it melts, and the part then quenched
Of course, as before, the two operations of hardening and temper-
ing are required.    Watch springs are tempered by the blazing-oif
of oil at a temperature of 5 70°, producing a dark blue.
*•   Hardening in Oil.—This may be looked on as a species of
tempering without preliminary  hardening.    It is of great value
when dealing with articles having very large surface, and which
could not be heated to an even colour by the methods previously
mentioned.   Only   one  degree of hardness  can,  however,   be
obtained, that corresponding to a dark straw colour in the table.
A pan of oil being provided of sufficient capacity, the article,
tieated to a dull red, is plunged into the oil; and the softer result
when compared with water hardening is no doubt due to slower
cooling.   (See Appendices I. and 11.,pp. 749 and 803.)

Gun cores axe cooled in oil, to enable them to withstand the
wear due to the shell, and also to increase the strength of the steel.
Thus, in some experiments at the Terre Noire works, four speci-
mens of steel were heated and cooled in oil, and it was found
that whereas the average breaking stress per square inch was
35-29 tons before the operation, it had afterwards increased to
51-23 tons.

It should finally be noticed that much care is required in
tempering—care not to overheat in the first operation; care not
to warp the tool in cooling; care not to crack the tool at the
ifater level. Some tools will harden best in a saturated solution
of salt, others in. a stream of running water. Generally it is wise
to move the tool well up and down during cooling. Hardened
steel may be compared to ^lass, annealed steel to lead, and

Steam Hammer for Forge.                    129

tempered steel to whalebone. Our process then when tempering
by the aid of water is to raise the steel to c glass/ and then lower
it gently to ' whalebone.' Hardening in oil gives the c whaleboneJ
without passing through the ' glass' stage.


We shall now pass on to describe the turning out of very
heavy forgings, which include all articles too ponderous for smith's
work, and which are consequently made in the forge under a very
heavy Steam-hammer. Fig. 124 is a drawing of a hammer
suitable for general forge work, such as we are about to consider,
but, of course, extra large forgings would require special-sized

The hammer in Fig. 124, Plate IT., has a falling weight of
five tons. After the careful account of the smith's harnmer there
will be -very little to say here by way of description. As before,
the outer valve is for the purpose of admitting steam (being
opened by a screw acting at the end of a lever), while the inner
valve controls the direction of flow, the exhaust passing upward.
The long hand-lever serves to move the distribution valve, and
the self-acting aim between it and the valve reverses the latter
as soon as the arm is moved by the tup on its upward travel.

The Furnace used hy the forgeman is very similar to that
shewn in Fig. 85. It is there called a Puddling Furnace, and
indeed * blooms' are to be made for heavy forgings just as in the
•case of puddling, the only difference being that they are built
from scrap iron instead of white pig. A pile of scrap iron is
heaped on a rough wooden tray, and is then put into the furnace.
Several of these piles being so placed and heated sufficiently, they
are then found stuck together. Withdrawing them, thus adhering, •
by means of very large tongs having a balance-weight on the
handle end, and supported at the middle by a crane, the blooms
are put under the hammer and well beaten together to form slabs.
It will be these slabs that we shall use to build up our forgings.
Fig, 125 shews an arrangement of furnace and cranes for heavy

First we shall consider, in detail, the forging of a Double-
throw crank shaft of large size, the finished form of which is






^ ^


Forging a Large Crank Shaft.                131

seen at A, Fig. 126. The forgernan always requires a staffer
'porter* to carry his forging, to which, for the time at least, the.
latter is welded. It is simply a long tapering bar B (Figs. 125
and 126)7 supported by a crane chain, and carried to and from
the furnace by the undermen, while the head forgeman directs
the hammerman, and applies the different tools to the work under
the hammer. The end of the porter is put in the furnace and
made to pick up, at a white heat, a few slabs which have been
previously placed there; putting them tinder the hammer they
are all thoroughly welded, and the round form of the first part
of the shaft obtained by swages similar to those of the smith, but
of suitable size. More slabs are added, and welded, until the
shaft is sufficiently long to take the first crank web. The web is
now built up by laying slabs upon it as at c (Fig. 126), the end
being put back in the furnace. Care must be taken in piling
these slabs, both now and always, that space be. left between them
by the placing of pieces of scrap, so as to enable them to take a
welding heat right through. Bringing the hot slabs back to the
hammer, they are welded by striking both at top and sides: and
so the process is repeated on both sides, a and I) (Fig. 126),
until the shaft has the form D (Fig. 1260). It is then set down
as at E. But the web is not yet finished. Heating again, it is
flattened out to the shape F, and slabs are again piled on and
welded to the body of the material, the process being repeated
as before for both sides of the web. The object of laying the
slabs on both sides of the web is to keep the direction of the fibre
such that the crank may be best suited to meet the stress pat
upon it. By this time the forging, being unbalanced, will be
difficult to turn round; but this is overcome by clamping four
arms dd on to the porter, these being turned by the strength of
two or four men as required. The web is now hammered at top,
bottom, and sides, to correct dimensions, the ragged end e chipped
off by means of a cutter, and the other end f cut down with the
same tool, the extra piece G (Fig. 126a) being worked by sets
until drawn out to receive more slabs. The shoulder & and the
piece G, are next finished to the round by means of swages, and
the building of the second web commences. This is carried out
in exactly the same manner as the first one, except that it must


Forging Steel.                                 133,                          |

be carefully built at  right angles ;   this   point, as   well   as that                          f

of the general straight ness of the shaft must be gauged with
square and straight-edge by the head forgeman, as the work

By this time then our forging has reached the condition Hy
and as the sketch A, Fig, 126, shews us a solid collar, for the
purpose of coupling to another shaft, we must add this portion.
Slabs are again piled up as at j, Fig. 1260, heated and welded,
until sufficient staff has been worked together to form a small collar
K, and then the whole collar can be finished either by the slab
method, or scarfed bars (L) can be wrapped round the shaft and
thoroughly welded Finally the collar can be chipped down at M
to the correct length, and cut off entirely at N. There only
remains the porter end o, which may he finished by taking off
the handles, and clamping them at the collar end, then putting
the porter through the furnace till it protrudes at the further doorf
and after heating cutting it off to the length shewn on the drawing.
The shaft is then set aside to cool.

Steel Shafts are forged from ingots (obtained by any of the
processes mentioned in Chap. Ill), and being thus treated from a.
solid block, differ in no sense, except size, from the example shewn
in Fig. 119. Some makers prefer, after flattening the ingot to the
thickness and height of the crank webs, to set down the central
portion of the shaft, forging each web in the same plane; and
afterwards, to turn one web at right- angles to the other by
twisting the shaft; but there can be little doubt that this is an
objectionable method, and should never be resorted to. A good
deal of care, in the case of steel, should be taken to get rid of the
blow-holes previously mentioned as existing in the ingots, and as
simple hammering is usually insufficient, cogging is the operation
performed, which consists in partly punching the steel while hot
immediately over any portion where honeycombing is suspected—
a sort of kneading, in fact

After the careful description of the crank-shaft forging, a share
explanation will suffice for the following articles—Piston-rod with
Crpss-head, and a Connecting-rod. Whenever such forgings are
made of wrought iron they are built up from scrap as in the case
of the shaft, such scrap consisting of ail kinds of wrought iron>


Forging Piston-rod.

especially the shearings of plates from the Boiler Yard, and this
being worked over and over again in the manner previously
described we naturally obtain a better quality of iron than that
which has been but once puddled. Another point to notice is
that the slabs should all he perfectly welded by good hammering
before the forging is actually formed to the required shape, for
much working after cutting to proper dimension will cause distor-
tion; while if, on the other hand, sufficient hammering is not
given to the slabs, cracks are sure to show after machining, and
the piece will be dangerously weak.

Fig. 127 will serve to shew the forging of the Piston-rod.
Its finished form is given at A, the cross-head being solid with the
rod, and having renewable ' slippers' of cast iron. Slabs are
piled on the porter to form the cross-head, as at B, first on one
side and then on the other; sufficient, if possible, to complete both

Forging Connecting-rod.


cross-head and rod. The shoulder of the butt is next knifed out
at c, and the rod drawn down and swaged as shewn at D, the
taper given at E, aad the ^vhole cut of to proper length. Finally,
the porter Is put through the furnace, as in the case of the crank
shaft, the clamps being transferred to E, and the butt end finished
by cutting off at F to the correct length.

The Connecting-rod in Fig, 128 is a little more difficult,

but no new principle is involved. A is the finished rod. Sufficient
material is first attached to the porter to make the forked end
and about half or more of the rod. This is shewn in progress at
B. It is next drawn down by sets and swages to the form c, and


Direction of Fibre.

more slabs are piled on to complete the rod and butt (see D and
E). No further description, will be needed to finish the forging,
as there only remains the cutting off of the butt end to correct
dimension, and the severing of porter from forging as in the
previous examples.

Want of further space compels us to close our chapter on
Forging, but sufficient examples have no doubt been given to
stimulate the student, who will now without difficulty be able to
construct other forgings for himself, albeit more complicated than
those already given. Of course different workmen have slightly
different ways of arranging their material, and no two will exactly
agree, but that forging will be the best one where the fibre is
disposed so as to meet in the best way the stress coming upon it.

Hydraulic Forging,—See Appendix //.,/. 803.
Cold Drawing.—See Appendix //.,/. 805.


THE pattern maker, moulder, and smith having supplied us
with rough castings and forgings, it is now necessary to finish
these articles truly before passing t^em on t6 the erector. After
marking or measuring-ofT, certain portions of nietal have to be                      I*

removed by hand or machine tools. The remainder of our work
will then consist of—Marking-off, or indicating the finished outline
by a boundary mark; Machining, or removing superfluous
material by automatic or semi-automatic machine power; and
Fitting^ which is the finishing of certain parts by hand power,
usually the chisel and file.

Machining has always tended to gradually usurp fitting by
hand, .and its advance is so rapid at present as entirely to take
the place of handwork for such articles as are to be repeated ; in
such instances manufacturers have special machines designed.
Even in unrepeated work a much larger quantity is done by
machine than hitherto, perhaps most of all by the extended use
of such tools as milling machines,

As so much depends on the perfection of a machine tool itself
{the workmen merely ' setting9 the work and arranging speeds),
a thorough knowledge of these machines is necessary, so as to
appreciate their capabilities and enable us to design work to suit

The next chapter has been reserved for the operations of
marking-orT, machining, and erecting, the present being devoted
to the Machines themselves, which may be classified as Lathes /
Planing, Shading and Slotting Machines; Baring and Drilling
Machines ; and Milling Machines.

Of course there are many general varieties of each class, and
each variety is again varied to suit special needs. Thus, as
regards drilling machines, most inland workshops are supplied

*m#jtot»''k*   €   §w«^Jf*#|
5    *f*',    i*    '^*ft               *f|   f*{

fcj?,   *«   *4|   |* Jst   ^f  ">^t^,# 9|
6,      ,   ,,»g     ***£    ^ H V»j -
'<,ff>    ^'^j,   ^ ./if      r v Hi ^ F   /f^!iftf |    <4   'I

a«J*|#    ^ f^V      '**'*    *'     *v   '   ^    '|*J4*?'*
it ^    /*  ^ -f^ff /f^i *a^|»   Ji"! |
'  '^    *   'l/f

l^l isr>f ^fi Jrt

U   ^MKt*-
^      ft!;

^ W* if

»i: ».¥?:?.!

Copying Principle.


feed) a plane surface;, and nothing could be more satisfactory.
If this be the completion of the cycle, as we suppose, then the
reciprocating tools, with lost back-stroke, must ultimately give

The Copying Principle is another great principle involved
in both hand and machine tools. All depend for their accuracy
on one or more carefully-prepared copies contained within the tool.
Thus in the carpenter's chisel the flat back is held against the
wood when paring, and constitutes the copy. The sole of a hand
plane serves the same purpose, its truth or otherwise beirjg copied
on the work, which may be proved by curving the sole, and thus
obtaining curved surfaces.

The copying principle is universal. Take the lathe : the bed
has a plane surface truly parallel to the line of centres, thus
enabling us to produce a true cylinder as our solid of revolution.
A second slide at right angles to the former gives us a
copy for use in (surfacing/ producing plane ends or right

The V grooves of the planing machine give accuracy along the
table, while the cross beam or slide ensures truth across it, and so
we obtain a true plane. The vertical slide and the two horizontal
cross slides are the copies in the slotting machine, while the
shaping machine hdis two copies supplied by the horizontal slides,
at right angles. Lastly, the milling machine has two slides, at
right angles and also horizontal.

As the truth or otherwise of these copies is transferred to the
work, it is of the utmost importance that they should be made
perfectly correct in the first instance.

The copying latfce and other duplex wood-working machines
are farther examples of the principle. (See Appendix II,,
p 812.)

/ Cutting Tools.—We will "now consider the shapes and
angles required for the tool itself. As a rule wood-working tools
act by wedging, or splitting-off the sharing; and the resistance is
tensile, with some bending. Our interest is with cutting tools for
rnetal, and Prof. R. H. Smith has shewn their action to be totally

Tlie diagram Fig. 129 represents the tool in action,    B is the


1 Cutting Action.

angle of relief or clearance angle, to keep the tool clear of the
work; A. the tutting angle, and c the tool angfe.

The point c requires great strength for metal tooling, and as
this makes A. very large, 'paring* cannot occur, but the material
will be 'crippled,' either by compression, shear, or a combination
of both. Sections parallel to F G will be in compression, and
those parallel to E G in shear, and it will be evident that along

of tool/


some section E F, the material will "be weakened to the greatest
extent ; here then the shaving breaks so ranch as to curve up the
face of the tool The direction of E r will depend on- the relative
values of the compression and shear strengths of the material,

Great heat is generated, due to  molecular resistance and

friction.    A lubricant of soap and water** is • used for ductile

materials like wrought iron, contained in a can placed above the

tooi-t»ox, and led to the topi point by a wire, down which it

* Or soda, if rusting is to be avoided.

Cutting Angles.                              14.1

trickles.    This cools the tool, and lessens  the friction between
tool and shaving.    For cast iron and brass these precautions are f
not needed.

There has been, up to the present, some diversity of language
regarding the angles A, B? and c (Fig. 129). Thus, in the planing tool,
A. has been termed the cutting angle, while in the lathe tool c has
been so called. Manifestly the first is the more reliable nomen-
clature; then c may be called the angle of the tool.

Their values were determined by Hart thus:—

For cast iron. .For wrought iron.   For brass.

A—Cutting angle   ......     54°     ......    55°     ......    66°

B—JR.elief angle.........      3°    ......      4°    ......      3°

C—Tool.angle   .........    51°    ......     51°    ......     63°

This supposed the least force of propulsion was required. But
if endurance of point he considered, a larger angle is usually-
given, as follows :-•-

For cast iron.  For wrought iron.   .For brass.

A.—Cutting angle  ......     70*    ......    65°    ......    80°

K—-Relief angle .........      3°    ......      4°    ......      3"

C~—Tool angle   .........    <>7&    ......    61°    ......    77°

In a lathe tool i* is termed the bottom rake, and j the top rakt,
while a third angle with top of tool, but on right, or left side, is
called side mke.

These angles will serve for any machine, and the shape of
tool and shank will be treated in its proper place.

The Screw-cutting Lathe.—Plate V. shews various
views of this, the oldest but most useful tool. The example is
the design of the Britannia Company, and has to in. centres, that
is, will accommodate work of 20 in. diameter (called in America a
20 in. lathe). 40 in. work can be turned by removing the gap
bridge A, which is bolted down and dowelled, so as to allow thtk
saddle to pass over it freely.

In all lathes the work is rotated, and the tool fixed in (usually)
a slide raV, which can he moved along the lathe bed, This ap-
pliance, the very foundation of machine-tool accuracy, was the
invention of Henry Maudslay. On account of the various
diameters to be turned, the angular velocity must be capable of


Revolutions of Lathe Mandrel.

variation, for the linear velocity at the surface of the work must j
be constant. Fig. 133 shews that if ab and ajbl are equal, the I
angle a^cb, must be greater than acb.

Let r=radius of Tvork in inches.

V — speed of cut in feet per minute.

N = revolutions per minute to produce V.




And as the''cutting,.sgejgds for roughing are, say:—

For brass...........................    60 feet per min.       ""J

For gun metal   ....................    50   „       ,,              S^

For cast iron.......................    40   „       ,,             <,

For wrought iron..................    40   „       ,,             ^

For steel..............................    30   „       „             .g

We have:—                                                                         ^

Revolutions per m. for brass......... =* 115-rad. in ins, :§»

„              „          gun metal ... =   95-        „        ^

„              „          cast iron   ... =    76-        „         -3

„              „          wrought iron =76-        „         *

,,              „          steel ......... =   57 ~       „        S

To effect this variation without altering the angular velocity of
the main shaft, cone pulleys and back gear are employed.

The cone pulley c is driven by a belt, from a like pulley on
the countershaft overhead, but tbe latter is reversed end for end,
so that its small diameter is opposite the large diameter on the
headstock. As the sum of driving and driven pulley diameters
is constant, the belt will fit any pair, and a change of velocity will
be effected, the highest being due to the smallest pulley on the
hejoVstock. (See Fig. 539,, p. 535.)

But as sufficient variation cannot thus be obtained we use the
jsgur "wheels knowiijas back gear. The mandrelD (Figs. 131 and
i34)*ls attached directly" to "the work by a driver. But the cone
pulley runs loose upon the mandrel. Deferring to Pig. 134, the
bolt E serves to connect the pulley with the wheel F, which is
keyed to D, and by sliding i radially outward till it engages
between lugs G an the pulley, F and c are united, and the
mandrel is driven cKrectly.

Slower speeds are obtained by releasing E, and allowing the



 k                              ^V
 5               ^

	1?                 J
 ^ g                     SW£W/tfG i   k!                  \SF/ t~)r i

	v      K    r»                                V£.t-,(*/L>/
	U      *?
	^ vj t
 • nl
 <C  o u

	• I1*
 S ^5 ^


fc 5>
	•oo «i* \          /k

	93    X                  SHEWSt ^                  /\fiSGU£-


	•      45                              j

	28                           ~

 "~"        A,





N]   154-










I    I ir     it it

, ^^

ii>. f w                   4,1



!/   •/*•                 ^|                 •>!              ^   ,t

>^'« li'^^ill*? f           iW f *if  *ift

^ *    *j"4*Mn4  ^ "A                       s*|l

^ <{     f                                                         *»||                                                                   %|

'* ^'Mn yw* ft#j i^| t#|

'-Vf »H  **>                   ^   ||                     |«>

-V   ^|i',*i     *?*^i»|^ti*|            ^  |t^f|, /%|


rt   ftf^*i*    4            »t||

f   **p ^ft       *

** ?4^^

l   \      |

* * '  f   j, '

1 >,    r '         |,      H

•    H    j': ' f    i

t  f > J

* t

/ - \

.'        «>l   '«,»*!

* ^i  \^,|n4
i! t



TAe Slide Rest.

Turning to the Slide-rest and its various feed motions, details
are shewn in Pigs. 130, 131, 141, and 14.2. x is the saddle,
having one rnovernent, that along the bed; Y is the middle slide,
moving atwss the bed; and the top slide z has a universal motion,
but by hand, being mounted on a circular table formed on Y ; and
thus a feed may be obtained at any angle by turning the upper
plate z and clamping the bolts a a.

The movement of x is'called traversing or sliding, and the
cross-movement of Y surfacing; these can be combined if re-
quired- The slide rest is actuated from the mandrel in two
distinct ways. The featfmg screw at the front of the lathe bed is
only used for screw cutting, and is thus preserved from wear at
other times. It is driven by * change wheels,* at the left end of
bed (Fig. 132). These can be changed, so that various rates of
rotation of screw can be effected, relative to that of the mandrel,
which'comparison fixes the fineness of thread cut on the work.
To facilitate the fixing of the wheels chosen, the intermediate stud
b is supported (Figs. 130 and 140) on a radial arm or quadrant 4
which can be clamped at various angles, the two wheels on b being
fastened together by keying to a loose sleeve d. The saddle and
leading screw are connected or disconnected by the two half
nuts e e (shewn apart in Fig. 141), which are brought together by
moving the handle downward along the dotted arc, when the
studs J/j carrying the nuts, are brought nearer the centre by
means of the curved grooves,

The slide rest is also worked from the back shaft h on the
opposite side of the bed, and the two feeds for surfacing and
sliding obtained. The shaft is driven from the mandrel by change
wheels (shewn dotted at £•, Fig. 131), the intermediates being
carried on the arm c. Some makers drive by belt, which may
slip if the machine Is being overworked, but there is no doubt
that wheels give a more definite feed. Passing to the connection
of shaft with saddle we refer to Figs. 141 and 130, A worm j\
having a feather kef, slides along the back shaft, being drawn
along by the saddle. The power passes through an intermediate
worm pinion 2 to the wheel 3, which, being keyed on spindle ^,
crossing the bed, rotates pinion 4 on the front side. This pinion,
gearing into wheel 5, turns the rack pinion 6, and the traverse is

1 ^v^CI'

f     J     4                     "    *f*»/*»  **»*#**,,.



^x^>2*Kfc£»i%3s«P^V^-<ix<-'V'.'.--- '/".'-. '*•-'f'.-'.'/ffS .'• . ;•    -•


^   *p>  «w*

V * ^ * *
l*~*4 7

? f
i f
* a



"* C  2,       **•&"********   -•ft*^

^ i ? * M I t * * * t ~ *" -

*  J * T ^ f I 5 * f f -       -  ^

F "t- # ^  ^  |   % ^      Z  *   \   -*•   "    .

-»        ^            vt        •*       *       •*      fpt       «£.                    •*

^             **    ~*                     4^             ,*A   S     ^

; - i f ? ^ * - ^ * t v

i s,- r , * " i * t -:

*,   *

« *






I      4 I'

>    1


4*   *%*fHlt>WJU    40f

^/14-      «      |H                  *||

yst, *^t   i

#    ft


*f >i/i:-^t^    v>   %/t

^  *"> «**#  k*'"'i
t vifiV/i* *^;

* i

t I

\ ;

^     J«*   f

' i         "     I J

,J*><   *f

y   jf * t7 < ** *

*4         .       %

-    .    f




Chucks.—Four examples of Whiton's chucks are shewn in
Pigs, 150, 151, 152, and 153. The Independent Chuck (Fig, 150)
Is really a dog chuck. The screws may be turned by a square
key at A, so far as to release the jaws altogether, which, being
reversed, as at n, serve to hold drills when boring stationary work,
or to take a longer grip on rotating work. Pig. 151 is a good
example of a concentric or l universal* scroll chuck. Applying a
key to the bevel pinion c, the wheel D is rotated, carrying on its
opposite surface what, on reference to front view, is seen to be a



spiral having three or four turns in its whole travel. The rotation
of this * scroll' moves the jaws nearer to or farther from the centre,
but equally, thus centreing and gripping the work at the same
time. Fig. 152 is a Lever Chuck having a scroll, but no gearing.
A tommy is inserted at E to turn the scroll F, while the rest of the
chuck GO is stationary, All these chucks are fastened to the
mandrel in the same manner, by bolting to a small face plate
screwed on the mandrel,

9 The Drill Chuck (Fig. 153) has the back portion H screwed on
tlie mandrel, vrhile the front part j carrying the jaws may be rotated;
the scrall is therefore stationary while the jaws are carried round
it Hand tightening is sufficient for small drills, the surface of j
being roughened for grip; greater tightness is obtained by using





'/«   *-

'/ V l   ^


Expanding Mandrel.


the key as shewn at K ; and, finally, the worm end of the spindle
is used, as at L, for large drills. As the worm only bears on j in
one direction, it is applied at the opposite hole M to release the

Chucks that are either independent^ universal^ or eccentric at
will, are also made, having combinations of the foregoing motions.

Expanding Mandrel.—There is still another plan of
support for work having a hole through its centre. It is fixed on
a mandrel (or spindle that can be centred in the lathe), of which
several sizes are kept, having a slight taper, one suitable for the
work being chosen; but a more expeditious tool is the expanding
mandrel in Fig. 154. The mandrel proper is coned at A, and
has three grooves of the same inclination as the cone, in which

ride keys so tapered that their outer surfaces form portions of one
cylinder. The mandrel is screwed with right and left hand
threads as shewn, and the advance of nut B will push the bars
cc up the incline, so expanding the cylinder to any diameter
within the limit of the tool 0 serves also as carrier for the work,


Lathe Tools.

and nut E on the right is for releasing the keys or for stead)*"
them. This tool is made by the Britannia Company.
>/ Cutting Tools for Lathes.—There are various opini1
on the proper shapes of these. Fig. 155 shews the most comii^
where A is the plan of a straight tool, B that of a right hand H
and c of a left hand tool; D being elevation for all three. Tb




i *f if?4 / **i
, >

jf*r|,^ t«i I
' liif^f

,(»/,' .

^              •*'•

S|ftl*5|, ?f:;;-f*f*^.u-sf-;

I**** ***

$;.                  -*j..   t- ;|-?«'i

,               4li   Ml^Vf*'*'. *    »»*:•:

{**»*J|    f ,t,    .«\f

I**!    jl||4      i^l*)'/^
jt/i    f|!^<
t«|                      14

^*|          "if   I

if         ,ttf||  MV-*|*

f              »j!?«    fi«

1^,'^   .^^  *
I*                ,»!*"  J*'*N

\i'i •#.*&•*

^   sff


I *   ;


The Break Latlie.                           159

The fast head-stock B has a large 'Cylindrical bearing at c, with
adjustable cap, while the pressure of the surfacing cut is taken by
the collars of the thrust bearing D. The face plate requires no
further description than that given for Fig. 143, except to say
that the jaw screws themselves take the grip, and that the jaw
boxes may be unbolted and the work attached directly to the
plate. The back of the plate has an annular spur wheel, driven
by a system of * treble gear/ We may turn the mandrel through
the four wheels E F G H in simple back gear; or directly, bolting H
to the cone pulley, and throwing out F and G by turning eccentric
bushes ; but if a slower speed be desired G is slid to the right, E
and F kept in gear, while wheel K and pinion M, keyed to the
'third shaft L, are moved to engage respectively with pinion j and
wheel N on face plate.

We have, therefore, three alternatives :—Direct driving without
gear; double-purchase gear, E into F, and G into H ; or treble-
purchase gear, E into F, j into K, and M into N. The last is
only required for large diameters of work.

The leading screw, lying within the lathe-bed at A, is driven,
by change wheels p, through shaft Q, and wheels R R at the right
end of bed. By removing the change wheels, the backshafts may
be put in gear, the power being taken from the belt x, passing
thence to the worm shaft u by spur wheels, and across to the rack
pinion, as in the previous lathe. The handle v will pull the lever
w, and clamp the leading screw nuts, while the traversing
motion may be reversed at x. The slide rest has the same
motions as have been described for Plate V., and the loose head-
stock needs no further description.

This machine tfs used:—(i) As a screw-cutting lathe with or
without gap; (2) as a face lathe. For the first the gap may be
varied by loosening the bolts which hold the bed Y to the founda-
tion z ] and by then applying a lever to boss K to turn a rack
pinion ^ £, so bring the bed nearer the face plate, the standard
d being also removed. The work would be supported between
the lathe eentres, and driven by a bolt in the face plate, or by
small drivers as usual

As a Face Lathe, the gap is widened; and* the upper parts
efg of the slide rest being removed, they are bolted on the

160                               Face Lathe.

standard at /*, which has a circular T groove to receive
clamping bolts, and admit of adjustment at various horizo:
angles, thus obtaining a traversing, surfacing, or oblique ft
The position of the standard is adjusted by loosening its four,
tion bolts, and applying a crowbar to the teeth jj. Feed is gr
by hand, but can be made automatic as a star feed, or by
overhead chain. By the former a star piece is keyed to the si
screw, and a projection on the face plate catches this at ev
revolution, giving it a small turn. By the second, a ch
attached to a crank pin on left end of mandrel, and passing ale
overhead pullies, actuates a ratchet on the slide screw, and gi
a small feed at each rotation.

If a face lathe be especially made for surfacing and very sh
traversing, the bed is placed across the line of centres. (,
Appendix II.,p. 808.)

The Boring Machine.—Figs. 161 and 162, Plate V]
represent two views of a horizontal boring machine designed
Messrs. Buckton & Co. As already mentioned, many bori
machines are made with vertical bars, as for marine engi
cylinders, the object being to balance the boring head, and p-
serve truth of surface; but if the bar b$ made very large as
rigid, as in the example, inaccuracy need not be feared. The
are two classes of horizontal machine: in one the work is fix
on a stationary bed, while the cutters travel, and in the other t
bed and work are advanced, the cutter bar having no longitudir.
movement. The latter is analogous to lathe boring. (^
Appendix //., /. 808.)

Referring to Figs. 130 and 131, Plate V., a cylindrical bar
placed between the lathe centres, and  driven by  catch plal
About half-way along this bar a longitudinal slot is made throu|
it, and a projecting cutter securely wedged therein.   The upp
slides y and z being removed, as well as the bearings q and
Fig. 131 (made separate for the purpose), the work is bolted ;
saddle x, by bolts placed in T grooves s s, and, as the bar rotatj
and gives the cut, the traversing feed advances the work to the toe

Boring machines are made on these principles, being, in fac
lathes specially designed for boring. The bed is made low, an
the fast head-stock high, the loose head-stock dispensed with, aq



/ BucAton, 8c Co.)

IS     15U     IS

•   ,3ffJ
-.. 'iJ.



» <    V It      '.- ,<

\t       <    >         ,'     i        ,'1r<,    '

f .41

' tt 14  »'„ »   i*   # /i y * ' >*


t 4 f

Cil^   il

Irfv-r                                 ^*

/I»VU'»<H'«; f^'J^ if »-

l^l«^l                  * /!!*)*          V4r^ *i'**fr' ^rf <|f?$Oi *v  f   •?»'     ii'*i^!|*'»

m *, 4^ p«*                          «l f *if      li  +*   **M*'  ^ ' fr*«* ^l !'Mi«*i

^*»f ,|            *% ft^ M^ntyswt-/ ^! v. *V^»  »n 4fl^t! *»> |^   iM*

,|            # #4                  *A  ll'r   tr^»   *'* ^*^ »,tws%*!r#l   '*>v   If   4fnr%

l^f,                                  *           ^t*%| il ?,rftrt !H 14* *<* I**Pf"t il^f

!»  |»^«   tl   »f^   ^   »'M  -4            *'/f%   *.f^                   IliT

4,                                           1%                    ii, fHc  ^»«*^ %»r  *, n

ilrf          ^»l *«!                          'I it         '**>


<            I-* A«                           Ptt          . f*i i*               H1 »*'^


Single-Geared Drilling Machine.

screw to a projecting arm^ and provided with slots for bolts, as in
the lathe face plate. The pillar g, which supports f, has rack
teeth turned upon it, so that the lifting apparatus may always
remain in gear, whatever the position of arm y;

The lifting gear is as follows: A spindle q turns a worm, gearing
into wheel/, which has on its axis a pinion engaging with teeth on
the pillar g. The handle k serves either for spindle q, or for
hand drilling when applied to the mandrel. Some machines have
a plain pillar, as in the next example. A very deep piece of work
is accommodated by bolting to the foot or bed, and swinging the
table out of the way.

In double-geared drills the countershaft is usually self-con-
tained, as at m ; and the pulley n is driven from main shaft; the
fast and loose pullies lying side by side, and the fork being moved
by handle /.

The Single-Geared Drilling Machine in Fig. 167
needs little further description. Back gear is dispensed with, and
the cone pulley A keyed to the mandrel. Hand drilling is pro-
vided for by the handle B on fly-wheel c. s is the hollow sleeve
driven by mitre wheels; and a feed screw at D takes the place of
the rack, being provided with a long key-way, while a key E (shewn 1
black) is fixed to spur wheel F, so that a feed may be obtained at
any height of drill spindle. The feed screw further passes
through a nut G, fixed to the casting H, and a rotation of F will
therefore raise or lower the screw; such Dotation being effected
by turning the hand-wheel on spindle j, ther latter carrying a pinion
K gearing into F. A socket L in drill spindle receives a cylindrical
projection on the screw, in which a race is turned; and a pin M,
passing through the spindle tangential to the race, allows the screw
to lift the spindle without affecting the rotation of the latter. In
the best machines the feed screw is a hollow sleeve.

The table and supporting arm are similar to the last example,
the lifting gear consisting of a handle N and worm p, worm-wheel
Q, and rack pinion, the rotation of the last lifting or lowering the
arm. The rack R is a sort of strut fitted between the top and
bottom collars of the pillar, but otherwise loose. If the table be
moved horizontally the rack is carried round the pillar, and
remains in gear with the pinion in all positions.




*         *   &"tF "t""""":' •"**-•"•

"-*-     »     II^A




Radial Drilling Mac/line.

The Radial Drilling Machine is most useful fu*
work not readily moved, and has been, since first designed.
in request for holes in steam cylinders or boilers.    The form
depends on the nature of the work, and is sometimes disf
with, and a trolley run' under the drill.    Then the radial ar»
be swung from a wall or roof stanchion.    The machine in 1* * -
has a stationary table, to the top or side of which the \* f
bolted ; and the tool is adjusted over the work (i) by an f* f   *
movement of arm B; (2) a traverse of the saddle c; (3) a *
fall of B.    The last is obtained by turning the spokes K,  **
through worm and wheel, rotate a pinion in rack F ; and i •
first both arm and pillar turn within the bed at x.    The m***
G is driven directly or by back gear, and mitre wheels H H tf *^ '
its motion to the spindle j, from which again the power i*«*  * - -
to the horizontal spindle L by mitre wheels K K.    As arm I -
rise or fall, K is supported by a bearing projecting through    - '*
in the hollow pillar D, and a feather key connects K and j.

The saddle c has bearings M M, and a sleeved mitre w I   *
drives the drill spindle P.    A bearing Q supports a short sj ** '
to which are keyed the mitre wheel R, spur wheel s, and
pulley T, from the last of which various rates of feed are obi ^
as usual; and power is given from L to s by a pinion u, Vh- *
having a feather key, follows the saddle, so as to keep ahv** > *

The drill pSwer passes therefore through five shafts, G, j *   ff
and P, but this is not considered complicated in view ***
advantages obtained.    The saddle is moved along the
turning the hand wheel v, which rotates a small pinion, gearki &
the rack w.   (See also Chap. VII., p. 303 ; and p. 1018.)

Drills.—Some forms are shown at Fig. 169, where A i\   ,#£
pointed and fits in taper hole in the spindle, the cotter a prevc- * v *
slip.    The method of sharpening is seen at b, c, and d, and nt **    **
ee increase endurance of point    B is a pin-drill, where vari;**
in diameter of circle cut is permitted by the movable cti t ?, ** *
wedged in the slot & a hole being first drilled in the wt*r£*
receive pin h.    Cutting angles have been previously discussc** I

The twist drill c, no doubt the very best form for accr*** ^
work, is much in favour.    A socket / fits the spindle, and


Slot-Drilling Machine.

the larger drills; a second socket k within the first is for medium
drills : and a third, within k, fits the smaller sizes. These are
carefully ground to fine taper, and are quite rigid.

The  Slot-Drilling  Machine (now metamorphosed into
the vertical milling machine) has a saddle carrying the drill spindle


as in Fig. 168, but arm B is made immovable. While rotating,
the. spin die also receives a traverse along slide B, taken from a
leading screw,, lying within B, in addition to the shaft L. The
•drill thus cuts out a circle that travels along a straight line, known
as a slot. Keyways and cotter holes are examples, and for such
work vertical and horizontal feeds are required. (See Appendix If,,
p. 810.)

Planing Machine.


The Planing Machine, as mentioned, is not strictly
economical, because the tool cuts in one direction only, and the
back stroke is wasted. To minimise this loss, and at the same
time reverse the strcfke without changing the continuous rotation
of main shaft, ingenious motions called quick returns have been

A large-sized planing machine is given in Plate IX., as made
by Messrs. Hulse & Co. The table, stiffened with ribs, and
having T grooves on its surface to receive clamping bolts, slides
in V grooves B B, made true and level, being the copies. Thus the
work travels, and the tool is fixed. The belt pulleys c, r», E are
loose on their shaft, but c and E are technically * fast' pullies,
because they drive the table, being fixed to pinions F and G. The
strap being on pulley c, pinion F engages with wheel H ; and
pinion j on the axis of H gears with K ; L in turn with M ; lastly
pinion N moves the rack P fastened to the table. A slow cutting
advance is thus obtained. At the end of the stroke the strap is
moved from c to E, and then K is driven directly from G, the rest
being as before. Dispensing with one pair of wheels we have
effected two objects—(i) a reversal of the stroke; (2) a quicker
rotation of N, or quick return to the table. (See App. Iff., p. 918.)

When at rest the strap is on loose pulley r>, and handle Q lies
at right angles to the bed. Being connected to strap fork through
levers R, s, u, inspection shews that Q moved to the right will give
the advance, and a reverse movement the return stroke. But,
once started, these motions are automatic, thus—Let the table be
returning leftward in Fig. 171, back stop x will at end of stroke
catch lever Y, and move it to the left, shifting the strap rapidly
from E to c, the advance pulley. If the table travel to th6 right,
stop z catches Y and puts the quick return in action. These
stops may be adjusted to give various lengths of stroke.

Two vertical standards a a bolted to the bed have slides on
their front edges, and are stayed by tube b. A cross slide c lies
across them, supported by vertical screws dd, passing through
long nuts at the back. On the slide are two saddles ee, carrying
other slides//; to give a vertical movement to the tool. Screws
d d are to adjust the cross slide to any desired height, after which
it is clamped by screws gg. A handle may turn shaft h, which is


K      *

>*  «   '   n  *        f     f
^ \   » i      £"   \   *»«*,*.

*  <   v  > *   ¥ • *• *•     •,


,      ..   I      .

»• *•:   *       -    , ,j


I  ./f   %,^     >$**i$*«M M   ty|   1  ^-fi        | If

"T ^f*^***' If/44 ^t*1'/ #*' lls*r «f^4^^, I^'
*M4 I4^^lf, |^,¥ff f I** ft '' .* *i, v V *
t          tf^lfi^         f /« t,*nt ', /

IW           tfflJffhi   !frf   *1   ^fL  **.%  t

§¥tf f  ^*   /ii**»|  M
H,             l^*f    If'*^!  ^#   '1,4^,   i


i 'I
*        i, '*     I  -  |

.ft-     I-    '"VI    5*H

, • -.   ,.. -1

!.'»*•         »'  ,| ,|

Shaping Machine.


tool altogether.    Special clamps 23 hold the tool,  having large

square holes to receive it, and the turning of the set 'screw serves

both to fix tool to clamp and clamp in tool box.

The tool itself is shaped as in Fig. 172, being so bent back as to

place the point nearly in a line with the hinge and prevent ' digging.'
The Shaping Machine is a planer with moving tool and

fixed work, having on this account some  advantage  for  small

articles ; for if a moving
table be employed, its
stroke must exceed the
length of work, so as to
leave space for the ac-
quisition of velocity in
such a heavy mass \
while the moving parts
in the shaping machine,
being much lighter, en-
able us to adjust the


absorbing less
work by friction.

The machine in Plate
X. is by Messrs. Smith
& Coventry. The tool
box A is fixed to a 'ram'
B, the sliding of which in
saddle c gives the cut. The saddle moves along the bed r> to give
the feed, and an arm E, cast upon it, supports a rocking lever
F, which actuates the ram through the rod H. The cone
pulley rotates right-handed, carrying on its shaft (which
extends the whole length of bed) a pinion K, giving wheel L a
left-handed rotation. L turns on a stud fixed to the arm E, and
carries a crank pin P, whose throw may be adjusted similarly to
that in Plate XI. A die on this pin slides in a slot M, formed in
the oscillating lever F. Referring to Fig. 174^, the uniform
rotation of L will give the ram a slow advance when travelling from
a to £, and a quick return from b to a, because -ab is a longer path
than ba, as shewn by the arrows; the proportion being 23 to 14



Shaping Machine.

in the example. The pinion K can slide on shaft z, and so keeps
always in gear with L, being driven by a feather key. Length of
stroke is adjusted by the position of P, but position of ram is
given by adjustment of the nut Q.

The table R, supporting the work (which is bolted to top or
side as found convenient), may be adjusted for height by the
handle s, which, by mitre gear T, rotates the screw within the nut
u, fixed to the bracket v. The horizontal position of the work
may be varied by moving v along the bed to the point required.

w is a mandrel upon which hollow cylindrical work may be
placed by removing the loose collar x, and gripping the work
between the cones, The bracket Y steadies the end of the mandrel.

Three feeds are required, each of which may be worked by
hand if desired The pinion 4 (Fig. 1740) drives wheel & carried
on a stud <?. An adjustable crank pin on g is connected to the
lever ^, which gives, through ratchet d} an intermittent rotation to
the spindle/. Upon this spindle is a worm gearing into a worm
wheel on the mandrel w, and thus a rotary feed is conveyed to
the mandrel. The latter may be used for such articles as lever
bosses, which are interrupted on one side by the lever arm, and
therefore unsuitable for lathe work. The second feed is a
horizontal motion of the saddle for work fixed on the table. A
crank pin /£, on the wheel L, is connected to the ratchet m, and
the motion transmitted by n to the wheel j>. p forms a nut
attached to the saddle, and as the screw q is fixed to the bed, it is
evident that a rotation of/ will advance the saddle along the screw.

The third feed is vertical, r is a bracket fixed to the saddle,
and s a rod sliding in r, as -well as in brackets //carried an the
ram. At each back stroke of the latter the tappet a/, on rod
r, is caught by the bracket r, and s is moved to the left, causing
the ratchet gear x to turn the screw y, and give a small vertical
advance to the tool box. When the ram reaches the end of the
advance stroke the tappet z in turn catches ^, moving s back to its
original position. The head A can be fixed at an angle to the
vertical by unclamping bolts 2 2, and refking, vrhen the last-
mentioned ffeed becomes angular; and the position of the tappets
may also be varied.

In addition to the abore, a fourth movement, enabling us to
fix the tool box at an angle while preserving the -vertical feed, is




to ft" 173.


Slotting Machine.                             17 3

obtained by means of the worm spindle /, provided with a handle,
and worm gearing into the segment f, which is pivoted at 3. We
may thus shape a corner or give a feed (by hand) for a concave
surface. The front of tool box is provided with the usual flap to
relieve the tool during the return stroke, and the tool itself takes
the same shape as that described for the planing machine.

The Slotting Machine is probably the least economical
of machine tools. While the planing machine takes simple
horizontal cuts, and the shaping machine tools cylindrical work
lying horizontally, the slotting machine is for the production of
vertical cylindrical and plane services. Though working at a
disadvantage in having to lift a heavy ram, this machine has
served a purpose, and is still used to a large extent. Smaller
work can generally be accommodated in a shaping machine, but
the slotting machine is used for heavier work, and is made more

Plate XI. "represents one of these machines, as made by Sir
J. Whitworth & Co. Power being given to the cone A, it may be
passed directly to the mandrel B, or through the back gear at c,
the back shaft being moved to the left or right, or (Fig. 175) to
put the wheels in or out of gear respectively. The power is further
taken from the mandrel to the ram through the medium of a quick
return motion. Looking at the front of the ram, and keeping our
attention on both views, the spur wheel E is driven by the pinion
F, and the motion transmitted to the crank disc G by pin H. The
spur wheel turns on the boss j, and the crank disc in K, their
centres being i| inches apart horizontally. Referring to Fig. 177,
if the spur wheel rotate uniformly it will pass through 10 divisions
while bringing the pin from H to Hly but through only 7 divisions
from H! to H, and the advance will bear the proportion of 10 to
the return 7. As some sliding takes place between pin H and
disc o, a die is provided. The rod L connects the crank disc
with the rarn M, and there are two adjustments; one at N to fix
the height of the ram; the other at p, where the rotation of two
screws is made to move the pin and regulate the throw of the
crank. A brake block Q, bearing on the crank disc, may be
tightened by screwing up the wedge R, and serves to fix the ram
in positions where it might fall on account of its weight

There are three feed motions, all taken from cam s, the


Milling' Machine.

Figs. 175 and 178, which is connected by a shaft with the din
At every rotation of the cam a vibration is given to the lev*
which is connected to the lever u (Fig. 179), carrying a rat
pawl, and a partial rotation of shaft v (Fig. 175) thus obtai
Both levers are provided with slots to adjust the amount of fe

The table w to support the work, is circular in form, and
worm teeth on its lower rim.    It is mounted on two slides x
Y, which are again supported on the bed slide z.    The sha
turns the bed screw g through the wheels e and f, giving a lo
tudinal feed, useful for cotter holes and such like.    Putting/
of gear by sliding, a cross feed is effected by wheels a and d,
former taking its motion from v by mitre gear, and the la
being fixed  on the cross  slide screw h,  so  that Y would
stationary and x would traverse.    The third feed is a rotatiot
the table obtained by the worm gearing above mentioned;
wheel d being slid out of gear, and b put in, the worm shaft ^
rotated, and its motion transmitted to wheel k, cast on the tal
This motion is analogous to that of the shaping machine mand

It has been customary to attach the tool directly to the n
and let the point scrape on the work during its return, giv
useless friction and wear, but it is now recognised that a flap
advisable, and such a tool box has been shewn. A spring on i
front or counter-balance at the back is necessary to bring the t
back to its work, gravity not being otherwise employable.

The form of tool may be as for previous machines.

The   Milling  Machine, though in  its  present  form
recent introduction, has been known for a very long period; \
it was not till milling cutters or ' mills' were produced  m<
cheaply and correctly by emery grinders that the principle coij
be sufficiently extended.                                                         1

Cutter.—As already mentioned, a rotating cutter is employs
to which the work is fed, and this we shall first discuss. Fig. ij
represents a spiral mill for tooling flat surfaces. All these mi
are keyed to a mandrel or cutter spindle, which is either rotati
between centres, or fixed into the catch plate and only centred!
its opposite end. Fig. 182 shews a key-seating or groovii
cutter for cutting key ways or as a parting tool Being giwuj
both on circumference and sides, it becomes narrower at ea^
re-grinding, and therefore inaccurate. This can be avoided \

•• ^ ^"^^j^^^^g^m

", "  > f ^/j| ••• '••• ft



*r»* «/
S    ,CM*/v//vC



Milling Cutters.                           *75

the use of the expansible cutter in Fig. 183, which is divided, at
ab by a plane slightly inclined to that of the cutter, and has thin
discs inserted to preserve the normal width. If a & were at right
angles to the axis a strip of uncut material would be left on the
work, which is here obviated, besides which, various widths of
grooves may be cut. Further, if required, two mills may be
placed on one spindle, the teeth being interlocked, and a groove
of about twice the former width thereby cut, but it is important
that the mills "be of exactly the same diameter, obtained by
grinding them together on the same spindle. Fig. 184 shews a
pair of heading or twin mills for forming the sides of hexagon nuts
or other parallel work, the width being varied by the insertion of
suitable packing, In Fig. 185, A is a mill for grooving a screw
tap, B for fluting a rimer, and c an angular mill for cutting the
teeth of other mills. (See Appendix //".,,/. 8r i.)

When a grooving mill is allowed to cut on its side only, sayr
when fixed in a vertical machine, it is termed a face cutter, but
such an application is not desirable.

The steel or * blank ' to form the cutter is turned to correct
diameter while soft, and the teeth then cut. It is next tempered*
to a straw colour, and the edges are finished by grinding with a
small emery wheel of the same shape as the mill c, Fig. 185.
Great care must be taken to avoid cracking while hardening", but
distortion is now removed by grinding the hardened mill.

Fig. 186 represents a cutter for forming the teeth of spur
wheels by removing the interspaces, a is the relief angle or bottom
rake, a side rake being provided by cutting the profile in an arc
eccentric to that of the point path when rotating. Thus b is th e
centre for formation of the cutting tooth surface, while c is the
centre of rotation. Now d d and e e are curves struck from fr, aad
sections on each of these lines would be rectangular, but a
section on de must take the shape shown at./ because d± </* Is-
greater than el e± as seen in the end view. But as d e is the path
of the point d during the revolution of the cutter, clearance or
relief angle is therefore given at the side, and the cutter is said to-
be ' backedoff/ Of course this method can. only be used with
cutters of tapering profile; it enables us to preserve both form
aad width of cutting tool, however much is removed from the
face, and is art improrernent on the old cutter, which became


Milling Cutters.


narrower on re-grinding. The space between the teeth is to
admit an emery wheel for grinding the faces.

Angle of tooth, although important, is still rather in dispute,
principally because the same cutter, to avoid expense, is being
used for various materials—a wrong procedure, without doubt,
Probably some variation on the angles already given is necessary,
because of the higher speed of cut. Experience seems to suggest
the following:—

Cutting angle ..................    80° to tangent.

Angle of relief..................    10° to tangent.

Front rake  .....................    10° to radius.

giving a tool angle of 70°. Small mills are made with radial
teeth, corresponding to a cutting angle of 90°. A side rake of
10° should be given, and the teeth cut spirally or obliquely on a
finishing tool.

Speeds. —-There is still more variation in practice regarding
these. They can be considerably higher than for other tools,
because each tooth is in contact for only a small portion of the
revolution, and has ample time to cool. The result is the more
highly finished work that has brought milling into favour. The
following speeds give the result of experience, and are fairly

Milling Speeds in feet per minute ; and revolutions per minute, in
terms of radius (r)" of cutter.

For Steel      . ...........
	ROUGH i Ft. per M.
	NG CUT. Rev. perM.
 Ft. per M.
	SG CUT. Rev. per M.

	30 40 60 80
	51. r
 11± r
 JL5£ r 190 r
 !°1 r
 *i3 r 190 r 228 r

„   Wrought Iron  ......
„   Cast Iron ...........
„   Gun Metal    .........
,,   Brass...,. .............

Universal Milling Machine.         „          179

The Universal Milling Machine vtas of American design in the
first place, andotie of these useful machines is shewn in Plate XII.,
as made by Messrs. Tangye.

The mandrel A is driven from the cone pulley B, either directly
or through the back gear, the latter being thrown out by the
handle c, which turns eccentric bushes as usual. The mandrel
is of large diameter, for stiffness, and revolves in coned bearings
D D, the thrust when using a face cutter being taken by the steel
tail pin E. A strong overhanging bracket F carries a small head
H and centre G, to support an edge cutter, which centre is
roughly adjusted by unbolting H, and finely by unscrewing the
check nuts. The bracket is usually made round, and that form
has some advantages, but is not so steady. The mill is either
supported between centres, and driven from the catch plate; or
has a shank similar to that described for the drill sockets at
Fig, 169, when it is further steadied by the outer centre G; the
latter is the more common method. A twin mill is shewn in
position. Sometimes tools are fixed in the holes shewn in the
catch plate at j, which is thus transformed into a face cutter, but
the points must all be placed in the same vertical plane, so that
each may take its proper share of work.

A vertical slide K, having square edges for rigidity under
heavy cuts, supports a knee bracket L, which carries the table M,
and between L and M are two slides N and p, the first for longitu-
dinal, and the'second for cross traversing. These swivel on the
circular table Q, formed by their common surfaces, and P is made
of extra kngth in plan to steady the table, a detail often
neglected. A special point is the improved means of traversing
the table. This is often effected by telescopic shafts with universal
joints connected to the end of the table, and these sometimes act
at such bad angles that the joints in crossing centres cause a slight
dwell, which is reproduced on the work. This is avoided in the
machine illustrated. A small cone pulley R on the mandrel
dri-ves the lower pulley s, keyed to the worm shaft T, This shaft
carries a worm, gearing into a worm wheel g. A telescopic shaft
u is connected to the inside of the worm wheel by a universal
joint, and to the mitre wheels vw by a corresponding joint; these
convey the motion to the screw x, whidh gives a cross feed to the



Vertical Milling

•/f '

table. They are fixed in the centre of the swivelling table, and
will transmit the feed motion with steadiness, even when the
table is swivelled up to 45°, say for cutting spiral mills, twist
drills, &c. By moving the hand lever y the mitre wheel w may
be drawn out of gear, and the cross feed given by hand, if
desired, a catch z ensuring the contact of the wheels when in gear.
The longitudinal feed from screw a is rather a setting motion,
there being few cases where other tban a cross feed is desired.

The handle b is to raise or lower the table, which it does by
turning the screw e through the medium of the worm gear d.

Other forms of machine are Vertical Milling Machines and
Profiling Machines. In the former the cutter spindle is vertical,
and a circular feed, as well as traverse, is given to the table. The
latter is a smaller tool, where a vertical mill is traversed by a hand
lever so as to accommodate itself to intricate forms. Good lubrica-,
tion is necessary for all mills, and should be supplied under pressure
from a small pump. (See also pp. 752, 811, 1020, and 1025.)

CUT Tf/l

. /(PP.

Dividing Head. — When milling teeth of wheels, cutters, rimers,
&c., the work is supported in centres shewn in Fig. 189, which
are fastened to a small bed and bolted to the machine table.
The wheel to be cut is fixed on a mandrel, and held in position



(by T&M$iJt£s JWwcJwrub 7Joa& C$)

Dividing Centres.


by the carrier A, which is screwed on the right hand centre,
fixed in spindle B B. The spindle BB may be turned through any
desired angle by the worm c and wheel D. E is a steel drum
provided with small holes, representing various exact divisions of
its circumference, and the point F can enter any of these, so as to
set the spindle in the desired position. Knowing the number of
spaces in the wheel to be cut, or flutes to the rimer, dram E,
called a dividing plate, can be placed in each position in turn,
and a cut taken. The heads G and H can be either bolted
directly to the table, or packed to any convenient height, to
accommodate a larger piece of work; or H may be bolted to the




table, G packed, and the centres placed at an angle, as shewn by
dotted lines, useful for tapered work. This is obtained by
releasing the screws j and L, when centre K may dip, and B be
tilted between the cheeks of H. B may be further turned at any
angle up to the vertical, for milling cutters of various angles, and
E has a conical socket to hold the mandrel supporting the work.
Similar centres are used when milling spiral cutters or twist


Machine Vice.

drills, but then the spindle must be rotated gradually, by change
wheels connected with the feed.

The Machine Vice is a very useful appliance for Shaping,
Milling, and Drilling machines. It is shewn at Fig. 190, and
is bolted to the table of the machine, its object being the holding
of work too small to be fastened down directly, or to facilitate the
setting and re-setting of such work. A great desideratum is that
the latter should bed firmly on the surface of the vice, accom-
plished in the example by the bevelled jaw plates A A, which pull
the work down at the same time as it is gripped, by sliding on the
bevelled surface. The nut B can be rapidly changed to any
notch, and fine adjustment be given by applying a tommy to the
screw c. The jaw D has a cylindrical shank and plate F ; it can
therefore be set at any horizontal angle, and the screw c will still
bear upon it normally. B is also provided with lips at o, to
resist upward pull.


WHEN an engine or machine is first projected, a rough
'general* drawing is made by the draughtsman, in order to
determine the relation of the several parts; after which the
6 detailing' takes place, which consists in drawing out each piece
separately to a large scale, and at the same time classifying the
work—putting all the forgings upon one set of sheets, and the
castings upon others—so as to facilitate the distribution of the
parts to the various shops and avoid delay.

Detail Drawings are fully provided with dimensions, and
have red lines drawn round surfaces that need Fitting or Machining,
viz., such as are required to fit or work together; and the Pattern
Maker and Smith are thus enabled to decide where to leave extra
material. It is the business of the Marker-off to ' line out' the
rough work received from the above men; that is, indicate by a
boundary line the amount of material to be removed by the
Fitter or Machinist. The work is then finished and passed on
to the Erector, who carefully puts it together to form the com-
pleted machine.

The Marker-off's Tools.—A large plane table or
Surface flate is first required. This is shewn in Fig. 191, and
its size varies with the average work to be lined-out upon it—
from 4 ft. by 2 ft. up to 12 ft. by 4 ft. It is well ribbed under-
neath to prevent any possible distortion, and is planed very truly,
being better also if filed up and a little scraping done upon it.
The edges should be planed truly and adjacently at right angles,
sd that squares may be applied to them when necessary. Lastly,
the feet should stand upon a firm bed of concrete, and be
adjusted until the surface of the table is truly level, which often
assists the marking-off considerably.

V blocks, to support cylindrical work upon the table, are


Marker-off's  Tools.                               185

shewn at A, Fig. 192; and Cubical blocks are also provided,
of several sizes, but each of known depth and so figured. They
have their surfaces truly parallel, and are used to gain greater
height for the Scribing block, as well as for the purpose of
packing up the work (see B, "Fig. 192).

The Scribing Block, Fig. 193, is a most important tool.
It consists of an upright pillar A, fixed in a base B, which has
been truly scraped underneath. Upon A slides the head D,
which can be set to any height by tightening the nut H, a pointer
or scriber E being at the same time fixed at any convenient angle
by nut G. Most scribing blocks have no other adjustment, but
in that shewn there is a screw at F for further accuracy; here
the head c is first clamped, and D left free until finally adjusted
by the screw F, after which D is firmly tightened and the scribing
clone. The scriber has one point straight and the other curved,
the uses of these being shewn, where j can be made to 'scribe' a
horizontal line on the work by moving the block along the table,
and H may serve to ' feel? the height of certain other work. The
scriber is of steel, well-hardened, and must be kept sharp by
rubbing on an oilstone.

The Hand Scriber (Fig. 194) is to the marker-off what the
pencil is to the draughtsman. It is pointed at one end, and
hooked at the other for hanging to the pocket.

Compasses and Trammels must be provided for striking
^                   arcs of various radii, and as some pressure is required to make a

i                     sufficiently clear line on the work, both these tools should be

sufficiently rigid; the former being supplied, for this purpose,
with an arc and screw. Both tools are shewn at Fig. 195.

Accurate measuring Rules, with inches divided into eighths

and tenths; Squares large and small (3 in. to 3 ft); Straight

Edges of different lengths;  and Callipers, both for internal

|f                  , and external measurement, are all necessary tools; while if the

ir                    work is too large to mark-off on arable it should be levelled, and

'fi                   all lines be drawn by reference to an ideal horizontal or vertical

^                    plane,  necessitating the  use of either a Spirit Level or the

I                    Square and Plumb-Bob shewn at Fig. 196, the latter being

| /               the only tool in favour with the best workmen, as levels are

I                   known to get out of order so easily.


Fitter s Tools.

Of Centre Punches two are required, the larger for mark-
ing main centres only, and the smaller, or Dotting Punch, for the
purpose of making a scribed line more lasting and apparent, by
marking a series of punches or £ dots' along its length.

A light crane arm and Weston block is also of use when work
of large size is to be manipulated.

^ Fitter's Tools.—Most fitting is done at the bench, the
work being gripped in a vice, of which there are two principal
kinds, 'Leg' and 'Parallel/ The old-fashioned Leg Vice,
made of wrought iron with steel-faced jaws, is still considerably
used, because capable of withstanding a large amount of hard

ff-,f    3r*rf   As, .A*M«f, I**'

1 K;

l>, wiiii h $?;
?ar r,, 11
/ ». In «*f-

* ,M

M                                              ^|||                                                    |^»







•iL "

f        I

\         ^


Files.                                     189

the elevation, and B the plan view. The Flat Chisel, Fig. 201,
is used to true up surfaces previous to filing; and the Round-
nosed Chisel, Fig. 202, is for chipping out concave flutings;
but the last is more of a machinist's than a fitter's tool, the lathe-
man and driller both using it for c drawing ' a centre-punch mark
or countersink, which has been begun untruly, by chipping a
little to one side of the depression so as to alter the position
of the centre, -after which the drill or square-centre is again

The point of a flat chisel is ground symmetrically on each
side, and should enclose an angle about equal to that of the V
screw-thread, viz,, 55°, though a slightly smaller angle may be
used in finishing. After chipping, the surface must be further
trued by filing.

Files may be classified in two ways : (i) by the contour,
both in length and in section; (2) by the kind of cut and degree
of fineness. The length must also be stated, measured along
the edge, not including the tang. The cut may be double or
single, the latter being also called ' float' cut, but as this is prin-
cipally used for saw files, it will not be considered further. Longi-
tudinally, files may be parallel or blunt, and taper or pointed;
and in cross section they may be flat, three square or triangular,
half-round, round, and square. The fineness of cut is repre-
sented by the terms rough, middle, bastard, second-cut, smooth,
and dead-smooth, the last four only being required by the Fitter.
Safe-edge files are those left uncut on one narrow edge, to serve
in filing a surface near a corner, without destroying the truth of
that at right angles to it. Files are either machine or hand cut,
of which the latter are most in favour. It will be seen there-
fore, from the previous information, that a particular file may be
described something as follows:—' 12 in. hand, taper, flat, bastard,
double-cut, safe-edge file.' As the teeth only cut in one direction
the file is analogous to a planing tool.

Scrapers still further true up a surface left by the file or
machine tool. They are made from old files, by grinding off the
teeth and sharpening the edges, and have three principal shapes
as shewn in Fig. 203 : Half-round (A), useful in scraping a bear-
ing ; three-square (B), sharpened on the long edge for truing up


g. 2O3.


/   / '/

*'l^       *f^|


HiZ "*

	1 ......
	L. ,
	I ..... i ......

	* ^ *.»!* 4
	*,»          *«  *

	'I f*
	#"** *w^*"»
	s»»   MKMmiwwnti
	qMMMHM*!.     «r*|
	* ^m^trmm^ f!

	v^ .
	L"   -   .

	1. v fj<
	l"-i.      '
 1                   *'

	1 S «   «,                < '«     ,-      -

	"* 1
	* "                 *    '      S


	J . *•

	' *#' "

 J   "  ' '




and Dies for bolts or spindles, and taps for the nuts in which
these spindles are to fit. Taps are made in "sets of three to
each diameter of screw, and to cut V threads of * Whitworth '
pitch; that is, whose pitch per diameter agrees with those in the
table devised by Sir Joseph Whitworth, .and which is here

V Threaded Screws (Whitworth).

 in ins.
	Threads per inch.
	Dia.    ' Threads in ins.     per inch.
	Dia.    ! Threads in ins,    ;. per inch.
 in ins.
	Threads per inch.
 *H?       '

		i          8
	2i   \   4
	4        7
	4    i    4
	»s- !

	ti   .   7
	**    i    3£

	if        6
	3          3l
	»& ;

	4        6
	3i        3l

	if       5
	3*        3*

	if        5
	3f        3

	I 0
	.   ij        4i
	4          3

	.    9
	2    i  4
	4i         4

Very rarely are taps used beyond i^ ins. diameter, larger sizes
being screwed in the lathe. The set of three is shewn at Fig, 205,
and includes 'taper7 (A), 6middle' (B), and 'plug' taps (c).

These are made by forming in the lathe a perfect screw thread
upon a f blank/ and afterwards fluting to the section shewn
enlarged at E, so that when the tap is turned right-handed it has
a cutting angle of 90°, and a small relief or clearance angle,
removed with the file. Next, two-thirds of the length of the taper
tap, and one-third of the middle tap are turned off, after which
all are hardened as shewn at page 127. When in use the nut
must first be tapped, and the bolt afterwards screwed to fit it
After drilling to 'tapping size/ that is, to the diameter at the
bottom of the screw thread, the taper tap is first entered (while
the nut is held in the vice), and is turned round by a wrench D
applied to the square on the top. Only when turned right-
handed is the thread cut, a» will be seen on reference to E,

Stock and Dies.                            193

Fig. 205 ; and a left-handed turn will release the tool. When the
taper tap has done its work the middle tap is introduced in like
manner, carrying the operation a little further, and finally the plug
tap is passed through to give the finishing cut. After every
stroke forward, the workman releases the tool slightly, so as to
avoid undue pressure and perhaps breakage. (Seep. 1025.)

A. stock and dies is shewn at Fig. 206. A is the stock, pro-
vided with handles for turning, and B is an enlarged view of one
of the dies, having a thread upon it in reverse, and four cutting
surfaces at 90°, two to each direction of rotation: so that the
thread may be cut both on advance and return. The dies are
shewn in position in the stock A, being dropped in at e and slid
along : then tightened by a tommy applied to the screw d. The
bolt to be screwed is first turned to the outside diameter of
the tap, and then fixed in the vice. The dies are separated
slightly, the stock brought over the bolt as at c, and the screw
advanced. The stock is now rotated until the length of the bolt
is traversed; then, on reversing the motion, a slightly increased
pressure given to the dies; and so the bolt is re-traversed again
and again, until so cut into by the dies as to show a perfect
thread, and gauge to proper diameter, which may be proved by
trying upon it the already tapped nut, and any degree of tightness
obtained after such trial. At each stroke a slight backward
release is given as before, and oil may he used as a lubricant
Various sized dies may also be applied to the same stock.

For screws under a \ in. diameter the Screw Plate in Fig.
207 replaces the stock and dies, and only one tap is required in-
stead of three.

The pitch of a screw being measured lefigthwise from centre to
centre of the threads, let us unwind the latter, both at the top and
bottom of the V groove. The diagram in Fig. 20$ will shew the
result obtained in each case, and it will be clearly seen that the
angle at the bottom of the thread is larger than that at its top.
But the action of the dies, in cutting, is to first mark out the top
of the thread with that part of the die formed to finish the angle
at the bottom, and it follows that by the time the thread is
finished, there will be an unnecessary endlong play of the bolt in
the nut. These faults are somewhat avoided by the use of the



Grindstones and Emery  Wheels.                195

Whitworth Guide Screwing Stock in Fig. 209. Here there
are three dies, a being the * guide,' cut so that its ridges just fit
the bolt at first, and are made to mark out the correct angle for
the top of the thread, b and c are the cutting dies (gradually
advanced by the wedge bolt d\ and these ultimately give the
correct form for the bottom of the thread. But the only perfectly
true method of cutting a screw is by means of the lathe, where
the tool is fixed in the slide rest and the thread formed by the
gradual advance of the rest coupled with the rotative movement
of the work. (See Appendix //,, p. 815,)

Machinists' requirements, in addition to the tools men-
tioned in Chapter V. These consist principally of grinding and
sharpening tools.

The Grindstone, though banished from some shops in
favour of emery, is still so extensively used as to deserve mention.
It is shewn at Fig. 210, and the stone fits on a square spindle
having journals at the ends, lying in simple bearings. Large
washers are placed on each face of the stone and the nut a
tightens these. Fast and loose pullies are provided for driving
by power, and a shield b to prevent the water flying about, the
latter being a necessary lubricant, c is a rest for the work, placed
rather high up, and as close to the stone as possible, to avoid
accidents. The direction of rotation of the stone is shewn by the
arrow, and the speed is such as to give from 800 to 1000 ft. per
minute surface velocity. It is not advisable to actually run the
machine in water as this tends to soften the stone.

The Emery Grinder is seen at Fig. 211. Its bearings are
longer than those of the grindstone, and its peripheral velocity
much higher, being about 5000 ft. per minute. A plentiful
supply of water is required for tool sharpening, otherwise, with
most emery wheels, the temper would be drawn and the wheel
become glazedf The water is shewn in the figure as coming from
a vessel above the wheel, but is sometimes supplied under pres-
sure from a small pump. Glazing is caused by the cementing
material becoming softened by the heat produced in grinding,
though properly the cement should wear gradually and fall away
with the emery powder, A very useful form of emery grinder is
shewn in Fig. 212, suitable both for tool grinding on the larger


iv is t-Drill Grinder.


wheel, and fluting, &c., on the small wheel. It is, „,made, by
Messrs. Selig, Sonnenthal & Co., and a small attachment is
provided to carry the wheel when grinding milling cutters, the
latter being then held on the spindle of the machine, and the

wheel driven by catgut band Emery wheels may be used for
general grinding and removing of surplus material and thereby
save a large amount of fitting. (See p. 1030.)

Twist-Drill Grinder.—These are of various designs,the one
in Fig. 213 being made by Messrs. Selig, Sonnenthal & Co. The end
of a twist-drill would be conical in shape but for the clearance or
relief angle. The true surface becomes, accounting for clearance,
a cone having a helix for its base, and enclosing an angle of. 118°.
A section of this cone, then, made at right angles to one of the
slant sides, would give a curve deviating slightly from a hyperbola,
due to the clearance. We will now examine the method by which
this hyperbolic surface is ground in Messrs, Selig's machine.


Capstan Head Lathe.


First the drill is clamped in a V groove made in the support
A, and is held in the proper position by means of the plate B
placed at the front end of the groove. The support A rides
on a guide-arm c, which, in plan, is set at an angle of 59°,
or half the angle of the drill. This allows the surface of the
drill point to lie parallel to that of the emery wheel a The
hand-wheel E serves to bring the drill to the wheel, and F turns
a screw for the purpose of taking up various surfaces of the
wheel so as to produce equal wear. G is a fulcrum, supporting
a rocking arm, which, in turn, carries a horizontal arm H. One
end of H encloses the emery wheel spindle, and the other is pinned
to the rotating disc j. It follows, therefore, that if the disc j be
turned left-handed by taking hold of the handle K, the rocking
arm will deviate to the front, and the centre of the emery wheel
will describe the approximate hyperbola required to be ground off
the drill point, as shewn by the dotted lines in elevation. By
fixing the fulcrum G at slightly varying heights by means of the
hand lever L, it is possible to obtain sufficient variation in the drill
curve to suit various sizes of drills; and, as the driving strap is
changed in position, it is kept tight by the jockey pulley N
provided with a balance weight. When using the machine the
workman takes hold of the handles M and K, and pulls K towards
him, and after one surface of the drill has been ground the latter
is turned round in the V groove, and the opposite surface trued
up, B then serving to register the second position with the first.

The Capstan or Turret Head.—Although we were
supposed to have completed our descriptions of machine tools in
Chapter V., our work would be incomplete without an account of
this very important labour-saving appliance. The lathe in Fig. 214
is shewn supplied with both Capstan-Head Slide Rest, and Screw-
Copying apparatus, and is designed by Selig, Sonnenthal & Co.
A is the head, which is capable of holding six tools, to be used in
succession on th^work in the lathe. These are placed in position
by releasing a catch E, turning the head by hand, allowing catch E to
return to its place by means of a spring, and finally clamping the
rest firmly by means of the lever D ; all this occupying but a very
short space of time. Of course, it may often be necessary to use
both slides to put the tools in position, as will be seen, and the


Erectors Tools.

usual rack is provided for moving the saddle seme distance along
the bed.

Turning now to the screwing gear, j is a rest for the screwing
head, with screw for adjustment, and when in position for work is
held by the handle N. At the same time a lever L, provided with
a screwed die fitting in the threads of the screw M, is placed at the
other end of the shaft H, so that when the screwing tool is on the
work, L engages with the screw M, but if the rest j be lifted and
thrown back, L is at the same time released. When in operation
the screw M is rotated from the mandrel by gearing of 2 : i, so that
a screw is cut at c, having half the pitch of the copy and of
reverse hand, M being usually left-handed, and c right-handed.
Of course the shaft H is capable of longitudinal motion, and the
piece M, being hollow, can be removed, and another of different
pitch applied, while the die, usually made of copper or soft brass,
does not need special cutting, but will find its way into the threads
of the screw.

Lastly, the lathe is provided with a hollow mandrel, which is
very useful for small articles that can be cut from a continuous
bar. An example of such work is shewn in progress, being the
making of a small tap bolt A hexagonal bar is held in a con-
centric chuck, drawn forward to a convenient length, and the
roughing tool g first applied, traversing to the front for position.
The bolt being roughed down, is finished by the tool h, and has its
end rounded by /. Next the screwing is performed by bringing
over the tool m ; and, lastly, the chamfering and parting are done
respectively by the tools k and /. It will be, therefore, clear that
a great deal of time and labour may be saved by the use of such
a tool where articles have to be made in quantity. All bolts and ;
studs are turned at such a lathe. (See Appendices, pp. 814, 978, I
1040.)                                                                                        ;

* Erector's Tools,—These must include Lifting Tackle and a |
Portable Drilling tool. The latter is known as the Ratchet!
Brace, and is shewn at Fig. 215 in position for drilling a hole, j
The pillar A is clamped to the work, and carries an arm F, which |
can be set at various heights, to take the brace and drill a As j
the latter is ground to cut in one direction only (see d, page 168), !
the brace is made to enclose a ratchet wheel c fixed to the drill i

L     H

•-—-*•.....*"l c \

f     f.%   •**%*»,*   j ****,*% jjfc,                  *   *,    »      „*****,

1       if


7    !   <"     M j       f#  f      .    i    '    f   ,   \

Jacks and Lifting Tackle                    205

mechanical advantage. There is a very large loss by friction
(some 70%), but this resistance is useful as serving to sustain the
weight when the chain is released by the hand. (Seep. 1047.)

Jacks are useful where overhead support cannot easily be
obtained. Fig 217 shews a simple Bottle Jack, the 'bottle7
serving as a fixed nut in which the screw rises when turned by a
tommy bar; and Fig. 219 represents a more powerful Jack 'with
worm gear. Here the screw is prevented from rotating by the
jaw dy and is, therefore, raised by the rotation of the worm wheel
A, which acts as a nut. In the example a handle of 14 ins.
radius turns, by means of a worm, a worm wheel of 16 teeth,
enclosing a screw of ij in. pitch; and a weight of 10 tons is
thereby lifted. The lower jaw d is for loads that are near the
ground, and the jack may be traversed, when in .position, by the
ratchet arm c, applied to the screw b at either end.

The Hydraulic Jack is both very useful and very interesting,
and is shewn at Fig. 218. It has an upper and a lower jaw to
suit various work, and both are part of the cylinder A. B is a
reservoir in which is placed oil, or water and glycerine. The handle
being moved upward on the fulcrum c, the pump plunger D is
thereby raised, and the liquid enters the pump through the
suction valve E; on the down stroke it is forced through the
delivery valve F, and exerts a pressure behind the ram G, thus
lifting the cylinder A. The valves are c non-return/ being loaded ,
by springs, and the ram is packed by a cup leather. It being
required to lower, the screw-down valve H is released, and the
liquid runs back to the reservoir. Screw j is for filling the latter,
and K is an air hole to assist the pump suction. The power
obtained depends both on the leverage and on the ratio of the
areas of plunger and ram, and may be calculated in the same way
as for the hydraulic press, which is discussed at/. 736.

There are a few other small tools of use to the Erector. The
D Cramp A, Fig. 219^, is for temporarily fastening two pieces
of work together; and the Key Drift B for releasing keys when
fitting wheels upon shafts. The file c is provided with a special
handle, usually made from a bent bolt, to enable a very large
surface to be filed; and the Square Drift at Fig. 219^ is really
a Fitter's tool, being used to clean out square holes too small to be

2 Tans £&&*# Jack-



I f



ffLE   FOR   LA&GE

. b.

Chipping and Filing.                         209

drilled and slotted. A Lead Hammer, for use on finished
work; a Hack Saw; and an adjustable spanner are also
advisable. Round holes are clearied by the Parallel Rimer
in Fig. 231, and taper holes by means of a Taper Rimer
similarly constructed.

General Processes.—Chipping.—Although hand pro-
cesses cannot well be taught on paper, a general idea may yet be
obtained. We will consider ourselves provided with a cubical
block of metal, and that it is desired to remove a rather large
amount of material from one of the surfaces. We commence by
placing the block on the marking-off table, and, chalking the
edges, scribe a line round as shewn at Fig. 2190, to indicate the
layer to be removed. This done we place the work in the vice
and chip with flat chisel a chamfer along the edge of the block,
nearly down to the scribed line, as at B, and make this fairly
straight with a rough file. Now the cross-cut chisel is applied,
and with it the cross grooves are cut as at c, each groove being
tried with a straight edge, to make sure it is not carried too far
below the general surface. We are now in a position to com-
mence the removal of the strips that remain by means of the flat
chisel, constantly trying the work with the straight edge, until the
whole is as perfect as the chisel can make it. The position of
the workman and the angle of chisel are shewn in Fig. 220, and
practice only will shew the steepness of angle required for the
deep cut, and the shallower angle for the lighter cut

Filing.—The file is next applied, and the various * cuts'
used in order from bastard to smooth. True filing requires con-
siderable skill, the tendency to the production of a convex surface
being very great The back stroke needs no pressure, as the
teeth do not then cut; but during the forward stroke all possible


21 o                                 Scraping.

pressure is put on with both hands, and the file carefully guided
in a perfectly horizontal direction, the position of the hands
being shewn in Fig. 221. Comparatively narrow surfaces that
are not to be scraped are generally smoothed by 'draw-filing,3 the
file teeth being rubbed with chalk to compel the small particles
to drop out, and thus avoid the scratching of the work, and a
still further polish given by means of fine emery cloth wrapped
round the file. The position is shewn at Fig. 222. There is
some difference in the grip of the file upon various materials, it
being greatest on wrought iron or steel, and least on cast iron or
brass, so that a file may best be used when new upon brass, then
on cast iron, and finally on wrought iron or steel, for it will grip
the latter when worn on the former; but the reverse method
would not be feasible. During filing the surface should be
constantly tested with straight edge, and when finishing, a hand
surface plate, being slightly greased with oil and red ochre, will,
on application to the \vork, at once indicate the parts to be taken
down, The skin of a casting should always be removed, either
by chipping or by pickling in dilute acid, before applying the file,
otherwise the teeth would be at once dulled by such a hard sur/ace.

Scraping.—If the surface is to be further trued, recourse is
had to the scraper. We will assume that the tool B, Fig. 203, is
to be used. It is held in the hand, as shewn in Fig. 223, and
the portions to be removed are discovered by smearing a hand
surface plate with oil and red ochre and applying the plate to the
work. Patches of colour will be transferred to' the higher por-
tions of the surface, and when these have been scraped down the
work is cleaned again and once more tried, when the colour
patches will be found larger in number, but smaller in size and
more evenly dispersed. The operation is continued until further
accuracy is hindered by the grain of tlie material. Then we have
what is known in the workshop as a trm plane.

Originating a Surface Plate.—When a new surface plate
is required it is generally topied from a standard plate kept in the
•workshop, the method of the last paragraph being employed-
But if no such standard be at hand, or if the truth of our first
pkte be doubted, it is necessary to use three plates in order to
originate a true surface. These three plates are first planed truly

212                       Screiv-cutting in Lathe.

by machine, and next filed with a smooth file to obliterate the
tool marks. We will indicate the plates by the numbers in
Fig. 2230. First (r) and (2) are scraped and tried by the colour-
patch method, then (2) and (3), and, finally (3) and (i), the cycle
of operation being repeated until all fit together with great
accuracy. The reason for this method is shewn in the diagram.
Thus—(i) and (2) may happen to be convex and concave ; then
(2) and (3) would be made concave and convex. But if (3) and
(i) be now put together, the convexity (or concavity) of both will
be apparent, and may, of course, be corrected. But when all
three fit equally well they must clearly be equally true.

Although fitting processes are less performed now than hereto-
fore, yet all the best work is trued up by the last-described
methods, after it comes off the machine, for however perfect the
latter may be, there is always some little distortion caused by
clamping the work, which, though slight, must be removed if
great.accuracy be required. (See Appendix //,/. 814.)

Cutting a Screw in the Lathe.—This cannot be fully
discussed until velocity ratio of toothed gearing has been entered
on, but the practical considerations may be detailed. It will be
clear, from what has been said in Chapter V., that if the leading
screw be connected to the mandrel in such a way as to revolve at
the same rate, a tool of the shape shewn in Fig. 224 will cut a
screw groove on the spindle that has been centred in the lathe,
of the same pitch as the leading screw thread. If, on the other
hand, the mandrel were to rotate at twice the velocity of the
leading screw, a screw of half the pitch would be formed on the
work, or of twice the number of threads per inch. Summing up
then, the pitch obtained will depend on the relative velocities of
mandrel and leading screw, a proportionately quicker speed of
mandrel giving a finer thread, and a slower speed a coarser thread.
The consideration of the proper change wheels to be introduced
will be left for Part II., but we may here point out that when
both shafts turn in the same direction the screw produced will be
right-handed (viz., same as its leading screw), and when revolving
oppositewise a left-handed screw will "be the result (Seep. 484.)

The correct section of V thread, as adopted by Sir Joseph
Whitworth, is shewn at Fig. 225, one-sbfth of the theoretical






214                                 Gauges.

depth being rounded off at top and at bottom, and the angle being
55°. The rounding at the bottom is given by the tool in Fig. 224,
but that at the top, as well as the general finish, is obtained by
hand-chasing tools. These are seen at Fig. 226, where A is for
the spindle, and B is for chasing the nut; the first being held
transversely and the second longitudinally. They are both
carefully cut to correct section of thread. (See App. //,/» 815.)

Fixing a Stud.—Studs are used in places where bolts are
inadmissible, because the material cannot be drilled right through.
The stud hole being drilled and tapped, and the stud having been
turned and screwed so as to fit tightly in the stud hole, the former
is entered, and a stud box placed upon the opposite end, as in
Fig. 227. Outwardly this tool has the appearance of the box
key described on page 113; but is screwed internally to fit the
stud, and has a small plate of copper at the bottom of the
socket to avoid damaging the work. A wrench being applied
to the square, the whole is advanced until "stopped by the
plain portion on the stud, when the box may be removed by
a sharp back turn.

Cylindrical Gauges are of great value in securing accurate
work. They are shewn at Fig. 2270. B being termed a * plug/

and A a ' ringJ gauge. The first is used for testing the accuracy
of a socket, and the second that of a pin, and both are made to
such perfection that the tested pin would be found to fit in its
socket freely, but with no appreciable shake. There are cases
where the ring gauge cannot be applied, and then the 'horseshoe/

Details of Horizontal Engine.                   215

form is used instead (p. 750), combining both internal and ex-
ternal gauge. For interchangeable work high and low gauges are
required, varying in size by a very slight but known amount, and
the aim is to make the work lie somewhere between the two, so
that any pair of parts will then fit, and the ' play' between them
never be more or less than certain fixed values. (See^. 277 / also
Appendix L, $. 750.)

Details of Horizontal Engine.—Having fully described
machines, tools, and general operations, we shall now proceed
to apply the information obtained to enable us to take piece by
piece the various parts constituting a 20 Horse-power Non-
condensing Engine, with automatic expansion gear; and, having
received such parts in the rough condition from the Smith or
Moulder, to follow them through their various stages, until put
together by the Erector to form the complete work. That course
has been thought advisable in dealing with this, the most impor-
tant chapter in Part I, in order to avoid any risk of omitting a
good example; it being supposed that if a student could
thoroughly discuss the whole of this machine he might be con-
sidered reasonably capable of thinking out any new case that
might be placed before him. In order to avoid repetition we will
make a few premises.

The Marker-off either chalks or white-washes his work before
commencing, and obtains the height for his scriber point by
first marking the same on the block B, Fig. 192, and then setting
the point to this mark. He should know something of the allow-
ances made by Smith and Pattern-maker, which are usually J in.
all over machined surfaces, and in extreme cases ^ in. Bed
plates, for example, warp \ in. or even more, and special material
must be left on their seatings,

Machining is marked on drawings to indicate all tooled
surfaces; being shewn by red lines; but in our case a thick
dotted line will serve the same purpose, thus: «• ••« — «•«•.
Further, although such drawings are copiously and fully supplied
with dimensions, these will be omitted in our examples, the scale
being given instead. The sizes represented on the drawing are
known as * finished sizes/ and the allowance on machined parts is
left to the judgment of the Pattern-maker or Smith.

In drilling, there are at least three various sizes that a hole

216                         General Directions.

may be made, although all figured the same on the drawings,
Clearance size is for bolt holes, so that the bolt may drop in
freely; tapping size is that at the bottom of the screw-thread ; and
gauge size is divided into 'working fit7 and 'driving fit/ the first
having both pairs made to gauge, and the second having its socket
to gauge and the pin callipered to suit the plug gauge.

As regards the drawings; these are classified as previously
described, but we shall further give each article a number in
Roman letters; and in nearly all cases the drawing itself will be
indicated by the letter A, while the various operations take the
succeeding letters of the alphabet. At the close of the descrip-
tions a e general arrangement' or complete drawing of the engine              ^
will be given, and we shall thus have followed in nearly every *
particular the practical methods of the workshop. One sheet              I
is omitted, that representing a collection of all the bolts and
studs to be used on the engine. This has been thought unneces-
sary, as the capstan lathe has already been described where these
parts are tooled. It may further be mentioned that there are
always more ways than one of performing the various operations,
both as regards sequence and the tools used, and it may also
follow that each method is equally good; in many cases, too,
where the marking-table is mentioned in our descriptions, it might
be found more convenient to scribe the work while in the machine.

I. Regulator Lever (Fig. 228).—This must first be set
upon edge, on blocks, as at B, until level; and a centre line
he scribed all round it. The circles may be struck, just to see
if the stuff c holds up/ and the length of the handle marked
off from these. Now punch all the five centres. (A method of
centering with scribing block by laying the lever successively on
its four sides and scribing any convenient height is shewn at D,)
Lay the lever next on its side (c), and pack up until the centres
are quite horizontal, as measured with scriber. Then scribe the
centre line all round, and mark at the same time the thickness
of the bossesf and of the lever itself, as measured from this
centre. Next put in the lathe, and square-centre; then turn and
polish the handle. Remove and clamp to the table of a shaping
machine, so as to shape across the flat parts; then clamp on .the
lathe face plate as shewn at E, for the purpose of drilling the

Regulator Lever.


bosses and turning them.    Of course the boss must be carefully
centred on the plate, and the blocking must be exactly the same



.ftiq. 22(9.


jf. J^et^er ^tfor

thickness.    The drill point is placed against the centre of the
boss, and the loose headstock brought up to the other end of the


Bracket for Lever.

drill. The latter is then packed in the slide rest, and firmly
grasped, when the drilling may proceed. The hosses are surfaced
to the scribed mark, and to the shaped lever; the diameter being
gauged by callipers. When these are finished, the sides are
marked out as at F, and the lever next clamped on an f angle
plate ' placed on the table of a planing machine (see G), being
packed at such an inclination that the edge may be planed • and
four settings are of course necessary. The angle plate is an
appliance which will be found useful for a variety of purposes.
Now finish off the lever by draw-filing and ernery cloth. If the
work be too long to allow the bosses to be turned, the latter
are tooled as separate pieces, having a portion of the lever attached,
and are afterwards welded to the handle by the smith.

II. Brackets for Regulator Lever (Fig. 229).—First
centre at the ends as at B, and punch; then try in the lathe to see
if there is sufficient stuff at the middle. Turn the shank to
dimensions, gauging \vith callipers, and cut the screw thread in
the lathe at the same time ; the taper of the shank being obtained
by setting the top slide of the rest by the requisite amount, and
giving a hand feed to the tool. Polish while in the machine, with
file aad emery, all but the collar, which maybe left rough, because
it is to be afterwards cut; the diameter then being made equal
to that across the corners of the hexagon. Now remove from the
lathe, and, setting again on the table in the position B, line out
the flat cheeks of the fork, and shape or mill these. Upon the
tooled surface thus obtained further lining is performed as shewn.
at c, the centres being again placed exactly horizontal. Strike
the pin hole and punch. Drill the hole In machine vice to
gauge, and, bolting down to the centre of a slotting or vertical-
milling table i>, tool all round with hand and machine feed
Once more line out, this time for the fork slot as at E, and also
mark a circle for drilling, making sure that the line** is taken for
this, not b. Drill the hole last marked, and take out the rest of tlie
fork slot in the slotting machine, finishing by cutting the oblique
portion In the vice by chipping arid filing (F). It may here be
mentioned that all "bright work is held in trie vice between plates of
lead resting on the jaws, and called * vice clams.' Lastly, cut the
hexagon on the collar bf dividing out as at G, and filing off dae flats.

j^JJZ^^                                      1 off- W

fit- 229

H                  J


/<                          L

> Lt&uer    3^off.   W.L




Regulator Details.

III.   Pins and Washers for  Regulator  Lever (Fi

230).—Three of these are required, of various sizes, to be mac
to the drawing H. Centre for the lathe ; turn to gauge and polii
as at j. The washer is made from a piece of plate, by fir
drilling the hole, and afterwards turning the rim on a mandrel, ;
shewn at K. Then, the lever, bracket, valve spindle, and linl
are all fitted together; a broach or parallel rimer (Fig. 231), <
exact gauge passed through each set of holes to clear out tl
irregularities produced in drilling, the pins put into place, an

the split pins marked off------a groove being cut in the washer \

at L, to prevent turning and undue wear.

It is advisable to make all pins of steel that have to withstan
much wear, and their corresponding lever bosses, if of wrougl
iron or mild steel, should be case-hardened

IV.    Links for Regulator Lever (Fig. 232) are marke
on a piece of plate as at M, which has first been planed on a
four sides, then drilled, cut in two pieces, and bolted togethe
They are finished off by filing in the vice, though, if large, the
would be slotted round, or milled.    Polish with emery.

V.    Regulator Valve Spindle (Fig. 233).—Lay this o
its side, in V blocks, as at B; centre the ends, and scribe th
flats.    Then put in lathe and turn to exact diameter, at the sam
time cutting the screw.    Remove, and tool the flats in a shapin
machine.    Now mark off the eye, as at c, and punch the centra
drill the hole to gauge, and take off the outer material with vertia
milling cutter fitting the curve a.    Finish off in vice and polish.

VI.    Nut for Valve Spindle (Fig. 234).—Lay on tabi<
and line out for thickness, as at B ; plane or shape the flats; maii
off the hole, as at c; and, placing the nut in a concentric chudj
bore and screw in the lathe as at D, so as to fit the valve spind^
easily.                                                                                     j

VII.    Regulator Valve (Fig. 235).—After cleaning witj
rough file to remove fins, this has only to be machined on certaij
surfaces, as shewn by thick dotted lines on the drawing A.    M
the face must be reasonably true with the lugs, first find centre d
the latter, as at B, and square a line from the back surface a <j
having previously blocked the hole with a piece of hard wood
do this for both lugs.    To produce this centre line on to thj

Regulator Details.


Value, $ritritZte     toff.   JtCJ.

edge of the plate, the valve is supported, as at c, on the marking-
table, so that its back surface, our present guide, is level, and the


Regulator Valve Box.

just scribed vertical line put in contact, both at back and front, with
a line which has been marked on the table. Then set up this line
with square, as shewn, so as to mark the centre of the valve plate,
and measure off to right and left the width of the valve. Scribe
also the thickness of the plate all round. Now set in machine
vice to plane the top surface, as at D, with a front tool, and the
edges with a side tool, and be sure that the travel of the tool is
exactly coincident with the scribed lines. The valve is now re-
moved, and treated similarly for planing at right angles to the
former direction. For this the fork is blocked, as at E, and the
centre squared; next produced upward, as at F, and the width of
valve marked, then planed, as at G. There only now remains the
cutting out of the fork, which is lined by squaring and scribing,
as at H and j; then j is planed out, and H is finished by hand.
Finally scrape the valve surface very truly, as described in a pre-
vious paragraph. It' should be mentioned that when wood is
used to block or bridge a hole, and a centre required, it is ad-
visable to shape a small piece of tin or zinc, as shewn at K, to
receive the compass-point.

VIII. Regulator Valve-box, Cover, and Gland (Fig.
236).—Commence by bridging or 'spanning' the two end holes,
and striking the circle representing the diameter of the flanges, as
shewn at B ; measure also the length of the box over the flanges,
and mark this. The valve-box is now to be mounted on the face
plate of a lathe, and as the casting is "rather long, it must be sup-
ported by angle plates, as shewn at c, being tightly bolted between
them, as well as having one flange fastened to the face plate.
Having been carefully adjusted until central, it is turned on one
flange and surfaced; reversed, and. turned on the opposite flange.
Next place the box on a planing machine, as at D, making sure it
is both level and square, packing if necessary, and having scribed
the top seating and boss, measuring from flange centre, plane
these. Remove, and bolt to slotting machine in a similar manner,
as at E, and slot the front face, measuring the distance a in finish-
ing. It should be noticed that two tools are here, necessary,
cranked respectively to the right and left hand, as at F. Set out*
the bolt-holes in circular flange as at G, and drill with clearance
drill. Set level as at G, and, squaring up the centre line, join this

224                     Regulator Valve Spindle.

along the top, as at H Then, in position H, scribe and square
the centre d, measuring the distance c from a straight edge at b.
Strike the circle for the hole, drill this tapping size, and tap to
suit the Bracket No. II. To ensure rigidity, the bracket should
have been screwed rather f full/ and be now taken down with
stock and dies until a perfect fit in d. The port is to be marked
off, as at j, with square and straight edge, and measurements from
the square flange, and the edges are then to be chipped and filed
by hand, an operation involving some trouble.

The Cover is to be planed and drilled. Find the centre of
the gland seating as at B, Fig. 237, and mark also the centres at
each end; then draw a line across. Set the cover level on the
marking-table, as at c, and squaring up the centre d, scribe the
thickness of the plate, and mark its width. Set central, on the
planing machine, in a machine vice, or its equivalent, as at D and
E. Level the cover to the scribed lines, and plane the side sur-
faces, a a and bl>, as well as the edges. Remove, and square
across for the adjacent sides, as shewn at F, using the gland seat-
ing as a guide. Then set in the planing machine, and tool the
edges; finish the surfaces e e to the same level as a and b. Now
reverse the plate, and, setting level in the machine vice, as at G,
scribe the gland surface to measure correctly from the surface a b e,
and plane. The cover is next to be marked for drilling, which is
shewn at H, and the holes gg drilled to clearance size, h to
tapping size, while j must first be drilled for the smaller diameter,
and the stuffing-box afterwards taken put with a pin drill specially
cut, as at j.

The Gland is first drilled in the lathe (B, Fig. 238), which
may be done more truly by blocking up the hole with wood
through its entire length, and letting the drill take this out. The
front may be surfaced at the same time. Next place upon a
mandrel as at c, and turn down to dimensions. Remove, span
the centre, and mark off the gland face as at D ; then drill the
holes (clearance), and finish off the edge with dead smooth file
and fine emery. The cover holes are lastly to be marked off on
the box by tracing through from the cover, then drilled and
tapped.                                                                       *

IX. Valve Spindles: Main and Expansion (Fig. 239),







Iffff 1


- 9




1L_J C


ni a

'J                         226                             Eccentric Rods.

As with No. IL, these are first centred, and. scribed on the flat
cheeks (D); then turned and screwed, and shaped on the cheeks.
The hole is next struck as at E, drilled, and the outer curve milled ;
and the fork (F) is taken out last by drilling and slotting. Broach
right through, and turn the pin as was described for No. III.

The Nuts are best finished by putting a number of them
after  drilling  to tapping size, upon a  mandrel, which is then
placed between dividing centres on a milling machine, and milled
}                            by means of twin mills (see B).    They are to be turned axially

:                        through 6*0° at each operation, and must be afterwards tapped,

1                         and chucked in the lathe for facing and chamfering.

ij                                  X. Expansion Eccentric Rod (Fig. 240).—Centre this;

'I                           also mark the length between the shoulders, and square up the

thickness of the T end.    Turn to the requisite taper by * setting
over5 the loose headstock, as shewn in plan at J,  so that the
j                            front surface of the rod will then be parallel with the lathe bed.

The amount of set-over will, of course, be equal to half the
difference of the two end diameters.    Surface also the T" end.
Remove from the lathe, lay level as at K, and scribe the cheeks b.
(''                           Square and scribe the tee at a to dimensions, measuring from

\\ '                          centre, and strike also the bolt-holes.    Drill these to clearance

\ !                        size, and shape a and b.    Then mark out the eye as at L and

i                         mill this with a cutter having the proper curvatures.    The rod is

;                            long, but as the milling only requires it to sweep through a semi-

* «                          circle, there will be no serious difficulty if it be well clamped.

> .;                              XI. Main Eccentric Rod (Fig. 241) presents no difficulty

\   ',                       after the previous descriptions.   (See Appendix //.,/. 818.)

C                               XII.   Intermediate Valve Rod (Fig. 242).—-This also

%                            would be tooled by previous methods.    The manner of fastening

ij                            the pin is worthy of notice.    The bearing surfaces of the fork are

|                            but narrow, and it is unwise to allow movement at that place;

the die, on the contrary, has a good wide surface, so it is there
only that wear should  be allowed.    After the pin is put in
position, a parallel hole is drilled right through the fork, and
|   !                         enlarged with taper rimer, the pin for this hole being turned in

the lathe with an oblique hand feed. All them pins are of steel]
and all wearing surfaces are cast-hardened.

The Die is surfaced and bored in the lathe, and after wards


228                              Guide Bracket.

shaped to dimensions, leaving sufficient excess of width to allow
of accurate fitting to the Radius Link, after which it is case-
hardened and polished.

XIII.   Guide Bracket for Valve Rods (Fig. 243).—The
machining is  shewn  on  the  drawing  at  A.     Set the  bracket
vertical by trial with square as at B, and line out the base a.
Scribe also the line b all round the casting, and at the proper
height   from   a.     Lay  level   on   its   side   as  at  c;   find  by
measurement the height of the boss and that of the foot centre.
Scribe the thickness of boss.    Next shape to these lines as at D,
the boss with a front tool and the foot with a side tool.    It
should be noted here that a side tool ought never to be used if it
can be avoided, for there is a great twisting action thereby pro-
duced which is calculated to wrench the tool from its box; but a
good deal is sometimes sacrificed to save two settings on the
machine.    Re-scribe the line b from the marks left on the side of
the boss, and lay the bracket on its side as at p, packing until the
centres are level; then scribe the heights of the large holes on
both faces and strike the circles.    The casting being hollow, the
core-hole must be spanned as-at G, in order to strike the bolt-
holes, whose centres are found by scribing a horizontal line and
squaring a vertical one when in the position c, and then bisecting
the right angles obtained.    The bolt-holes are drilled as at H, but
the large holes are bored in the lathe, the bracket being clamped
to the face plate, and the latter provided with a balance weight.
This will be understood from B, where the face plate will be seen
dotted, and the bracket clamped in position for boring b.    In all
such cases it is necessary to first drill a hole large enough to
admit the boring tool.

The Bushes are bored in a chuck, and finished on a mandrel,
and afterwards driven into the bracket, a block of wood being
placed upon them* to receive the blow of the hammer;

The oilcup cover is drilled for the hinge-pin, and finished by
hand, and the oil-holes drilled, and countersunk slightly at the
top. An £" spiral channelj/^hpuld be chipped in each bush with
round-nosed chiseLto.,allow the oil to flow.

XIV.    Eccentric Sheave and Straps (Fig, 244).—The
castriron sheave will be taken first    It is of the solid, form, being



slipped on tp the shaft lengthwise. But there are many cases
where it is necessary that the sheave should be in halves for this
purpose, and the machining would-be then performed in a very
similar manner to the eccentric straps to be described, namely, by
bolting together the halves before turning. The drawing of the
sheave is shewn at A. Lay the casting level on the marking-table,
as at A!, and scribe the various thicknesses; span the hole, as at B,
and strike a circle for its diameter. Grip in the dog-chuck, as at c?
bore, and surface the projecting boss and the face of the sheave,
marking the diameter of the boss in the lathe. As the sheave
has to be chucked eccentrically, the face plate must, of course, be
balanced. Next reverse, and turn the opposite face of the
sheave, this time chucking centrally, as at D, and setting the
already turned boss close to the face plate. Lastly, the outer
circle is struck out, as at E, by re-bridging the centre, and marking
the exact eccentricity on the centre line at x, and the work is
then bolted to the face plate, as shewn at F, each portion of the
rim being measured in position, and carefully turned exactly to
dimension, because it must be a correct ' working fit' with the
strap. The key-way may be slotted out. (See Apf. /,/. 751.)

The Straps (drawn at a, Fig. 244) are first marked off, as at
AU and B! (Fig. 245), with the proper allowance for machining the
feet, and the two are then bolted down together to the planing
table, as at E^ The bolt holes are next scribed and squared, as
at DU the casting lying level on its side, and these are drilled, as at
Et. The thickness of the feet for the front strap being lined at
F!, and the stop for the bolt-head at H, these are cut out, the
first with pin drill, as at G (known as 'knifing' or lface-arboring'),
and the second with chisel and file. Now place face to face in
the vice, and broach the bolt-holes right through; then, having
turned the bolts to a good fit, fasten both straps together. Lay
the bolted straps level, as a,t j, scribe the width, and grip in a
dogchuck, as at K, to face both sides, setting carefully for each.
Leaving the work in the chuck, examine the outer rim for
centrality (for this cannot afterwards be turned), and mark off
the inner circle for boring, measuring with callipers and rule as
the work proceeds. Remove from the chuck, and scribe the
remaining sttrfaces—b c d, as at M, measuring from the turned

*l                 Jp            tl         *tll

*; M^*                }»                             * «i                                ^|

IjKMt'l,    ^fiftf          '«(Y*^rf       y|^                   WJ    *^»*ty

4^i      f/*ff      *.HMfV*^'      -^      f><^<^^         >>',*     f.''*^           ^|^l*/4


'f   *   \", s       jf f   *#,^   ^p tit,*,*, tr   ^'tl^       }/ • »

*?   '   ,         t         *"'    #       *       ,/"#        »        Jr   4        |,       f *fl         ^'"Viifjf        ^         Jf    f|

^ » f ^ / i 5 ^  f    j ;  .• f   » « # ^*      * r fr< t

'^ i*     ' !*               *      **   4 *    *         ' r'r    *• Jf' ** *    »  *2    {/   >}lf       f

^ , 4   r           ^ /   * * t  s ' f i I    f *;     -     f            ' 1 1


t ',            I I

, i      : -       I   '



^^ ^ i  >t" >jf|

____________ J
	!   '



	j t



i i


------- j




_^*^~ -

-                                                                                x ^

^    "




"' ' I
*   "   '

/ i   '       • t




XVII. Crank Shaft Bearing (Fig. 248). The bearing
is first'laid level on its side as at B, the centre line obtained,
and scribed -round. The seatings for the brasses are measured
and also scribed, after which the bearing is set up as at c (see
both views), and adjusted by line and square till plumb. The
foot is next lined, and the top of the bearing taken from this,
making sure that there is sufficient stuff left in the bush socket.
Now plane in turn the seating sides, the foot bottom, and the
top. Stand the casting again on the marking table as at D, and
find the centre of the socket. Square this up, as well as the
socket sides. Scribe the bottom of the socket, measuring from
the foot, and line the bearing centre all round. Square these
lines across as in plan at E, and mark them on the opposite side.
Find the centre of the set screw-holes, and measure the foot bolt-
holes from the vertical centre line. Plane out the bush socket—
the sides with a side tool, and the bottom with a front tool,
finishing with a flat tool, and the corners with the ' corner' tool
shewn at F. Drill the holes.

The Cap or 'keep' is set on edge to line the seatings and
scribe the two bolt-holes and oil-holes, as at G, being first, how-
ever, planed to thickness on its bottom surface. After planing
also the seatings, and drilling the holes, it is'placed in position
on the top of the bearings, and the bolt-holes marked through to
the latter. These are next drilled and tapped, and the studs put
in place.

The Brasses are shewn in Fig. 249. Being first laid on its
side, as at A, the large brass has its width marked and its lips
lined for thickness, and is then planed. The front and sides are
next lined out on*all surfaces to dimensions, measuring from the
planed surface, and trying for depth of stuff between the lips, the
brass being meanwhile packed with sides truly vertical, as at B.
The whole is now planed by clamping in the successive positions,
c, D, and B, so that every surface is done, either with a side or front
or knife tool, the depth of the middle surface being gauged from the
lip; and the small brass F is similarly treated. (SeeApp. /.,/. 752.)

The packing plates are next machined, and all is ready to put
together. The brasses and packing are to be carefully smooth-
filed and scraped until they bed perfectly into their places in the

- -,-d *




X       X
X       5







*i M/* **>*!    i   ,   ±'?" >t   /,   *, ^r^i   n
"f ^ r *   '"    ^8    < 11>  s   •*, * *  f ^'I| *   , ^ |

/  , /"  '«   '   i    *,«./!  *'4|







[ ,"





:.$                             242                              Connecting Rod.

shewn.     Trammel  between the  shoulders,  and square up the

vertical lines, as well as the end lines; measured from  these.

Next lay the rod flat as at c, and scribe the centre line round in a

similar manner to the last.    Punch centres, and place in lathe,

testing with chalk, and square-centreing.    Now set the poppet

head over by half the diameter difference, and turn the taper

portion up to the shoulder radii, as shewn at j, Fig. 240.    Set

\                                  the poppet head true again, and surface the ends of the rod, also

J\a                                  the shoulders up to the radii.    These last require very careful

turning.    They are to be roughed out by means of a combination

of surfacing and  traverse  feed, and   semi-finished  by a broad

tool  ground  to the  curve,  the  position of which is gradually

changed by turning round the top rest until the whole curve is

gone over piece by piece.    The last finish is given by hand with

the same tool very sharply ground.    Of course the work must be

continually tried by means of a sheet iron copy called a ' tern-

,                                    plate/ shewn at r> and E, the lathe being stopped at each trial;

n"                                  and the outer curve of the solid end is to be finished in like

jj     ,                       manner.

<j     j                            Remove from the lathe, finished, but not polished, and lay on

f                            the surface plate as at F, packing till level.    Scribe the thickness

of the butt and solid end, then fasten, as at G, across a shaping

^                       machine having two tables, and shape.    Similarly also for the

depth.   Return to  the marking table.    Scribe the centre line

;                       afresh, and plot out the square hole in solid end as at j; do this

on both sides, and well dot all round it.    This may now be cut

out in one of two ways—(i) a hole may be drilled large enough

i                        o pass a slotting tool, by twist drill and pin drill, and the rest of

j                       the work "done by slotting; (2) a probably better method, is to

J                       take out all round by means of slot-drilling tool, drilling, say, a

;                       quarter of an inch down, traversing all round as at K, then a little

i                       further down, and so on till the hole is completely cut, finishing

J                       the sides with a milling cutter and the corners with a corner tool

;                       There is then very little work left for the file.

Now mark off the bolt-holes at L, on both sides of the solid
end, together with the oil-hole, and the cotter-hole at M. Drill
the bolt-hole from each side, broach through, and countersink
the oil-hole. Take out the cotter-hole in slot-drilling machine

Cotter and Strap.                            243

as above described, until cut right through, when there ought to
be little finishing with file. Drill and tap for set screw in butt
end, return the rod to the lathe for polishing, chip off the
centreing pieces, draw-file, and polish the ends.

The Strap (Fig. 253), being forged fairly to shape, is first
scribed to thickness, and planed. A sheet-iron template is next
provided of the form shewn at N, Fig. 253, which is placed on
the forging and the form traced. Finish the contour with
vertical mill or slotting-tool, clamping the work as at o for the
outer and as at P for the inner tooling. The oil-cup is next
marked off as at Q, and the strap clamped to an angle-plate as at
r< for turning, boring, and drilling. At the same time the screw
is chased for the oil-cup cover. Lastly, line out the cotter hole
as at s, and slot drill by blocking up in .the machine vice as at T,
and, on removal, draw-file and polish.

The Cotter u is first planed to thickness from good steel,
and then marked off to length and width. Both edges are then
planed to the marked lines, and the rest finished very exactly by
file, with the aid of the gauge template v, great care being taken
regarding the thickness.

The Gib w is similarly marked out, and the sloping edge
planed. The channel is then removed with a shaping tool,
several gibs being bolted together for economy, and the rest
finished very carefully with the file.

The Large Brasses are marked off and planed in the same
way as were those for the bearing, Fig. 249, and are then bolted
down very firmly to the boring table as at x, with liners between
to represent the draw of the cotter, and with bolts lying close up
to the outer surfaces. See their faces are set at right angles to
the boring bar, which is inserted as before, and the work traversed
into position. Bore right through, and finish the radii with a
specially ground tool, as at Y.

The Small Brasses are shewn at d and z. They are
planed as before, with the exception of the sloping side, which
requires a new setting, as shewn at e} and is planed with a side
tool. The 'ring' faces must also be left untouched, these being
turned at the same time as the hole is bored, which is done by
bolting the two brasses together, with a wedge between for the

Crosshead.                                   245

slant edge, and a liner to represent adjustment allowance, and
the whole chucked in the lathe, as at / Two settings are of
course necessary. (See Appendix /., p. 752.)

The "Wedge is now shaped to dimension, but not drilled;
the wedge bolt and set screw for the large end both prepared.
The oil-cup cover is then turned and screwed in a concentric
chuck, and milled on the hexagonal faces with a horizontal tool
by placing the work on a dividing plate. All is now ready to put
together. For the small end, fit the brasses in place by smooth
filing and scraping; fit the wedge, and mark off the hole for bolt
by scribing through the rod end. Remove wedge for drilling and
tapping, then replace. For the large end, the gib and cotter are
first carefully fitted to their holes separately; then the brasses are
fitted to the strap, and the latter to the butt end. Place all
together, and file the cotter till it enters the proper amount; then
mark off the split-pin hole and drill. Once more replace all
parts, and the connecting rod is complete.

XXII. Crosshead (Fig. 254).—Centre the forging, as at
B, and line the width across the cheeks; then turn the side and
end, and shape the flats. Lay now upon the marking table, as at
c (see both views), and scribe the horizontal centre line. Find
the centre for the gudgeon hole, as at a and £, measuring from a
straight edge, and test also with dividers; erect this line with
square, and strike the circle on both sides, also the contour of the
boss. Chuck in the lathe, as at L, and bore the hole, first
drilling to admit the boring tool. Remove from the dogs, and
insert next in a large bell-chuck, as at D, the exact position
being found by placing the work between the lathe centres; after-
wards firmly tightening up the screws, as shewn. First drill the
hole as large as allowable, the tool being centred, as at F, and
clamped in the slide rest; and next bore the taper, as at E, by
turning the top slide of the rest to the required angle, the feed
being obtained by a small pulley on the screw, driven from the
countershaft by catgut band. The hole is tested for diameter with
callipers, and the angle of the rest noticed before removing (this
being afterwards required for the Piston Rod). Now place the
crosshead on the mandrel of a shaping machine, as at G, and sbape
all round up to the return curve, the latter being tooled with a

f# in*   / *j^             */ ***/%




t :   '  "   /       //I %'

<*~,:   ,      ^u^fl

3   V

f »

Piston Rod and Piston.                       247

concave feed (mentioned in Chap. V.), and the flat portion with
horizontal feed. Take again to marking table, block level, as at
H, and scribe both fork and slot-hole, measuring from shoulder.
Drill and slot the fork, and slot-drill the cotter hole. Finally, slot
out the key way to suit the gudgeon; prepare a cotter, and take
out the taper in the hole with round file; then draw-file and
polish. (See Appendix /.,/. 752.)

XXIII.    Piston Rod (Fig.  255).—Centre on V blocks,
and set in the lathe.    Then traverse all over the work to the
diameter at c.    Mark off the various lengths a, b, c, d, and put a
centre pop at each place.    Turn down d to the larger diameter,
f                   and take down the taper at b and d by setting the slide rest, as at

E (Fig. 254). and it should be noticed that if the rest be placed
at the same angle both for rod and hole, the one is bound to
accurately fit within the other. Turn down at a to screwing size,
and chase ; then finish and polish the whole.

The Nut may be turned, bored, and screwed in a chuck, and
the hexagon milled. Lastly, the rod is fitted into the crosshead,
and the cotter hole marked through to the latter, then slot-drilled,
and finished with file.

I                       XXIV.    Piston (Fig. 256).—This is to be turned on the

f<                    rim, and bored to fit the piston rod.   The latter operation is clone

|                   at B, and the former upon a taper mandrel at c, the grooves being

turned at the same time to exact gauge, so as to fit the rings as
^                    truly as is consistent with freedom.    The plug holes, ^, left during

casting (see B, p. 30) are to be drilled and tapped, centres unim-
portant, the plugs being made from a round bar, screwed in the
;                    lathe and parted off to length.    They may be an easy fit in the

I                   holes, but must be painted with sal-ammoniac, so as to form a

rust joint.

The rings are rolled from J-inch brass bar, being received at

/                   the works ready formed, sprung out to a somewhat larger diameter

I                    than the cylinder.    The joint is shewn at a (Fig. 256), and should

1                   be as nearly as possible closed when the piston is in place.

j                         XXV.    Radius Link (Fig. 257).—The forging should be

I                    fairly to shape, being made to template.    First line to thickness,

»                    and plane.    Make a template exactly to drawing A, with the ex-

?;                    ception of the holes, which consist of quarter circles, as at B, with

VMi'»'» #**f     ^Xrf<lfI'"*1**,
.„.*,*«.,   X'



* *

I if


,f                   V;


Radius Link.                                249

a little piece filed out at the centre to admit the scriber. Lay
this template on the work, and trace out. Then drill the holes,
which are to be broached when all the parts are put together.
Remove all the outside material with a vertical milling tool having
a radius equal to that of the return curves, as shewn at c. The
inner slot may be cut out by one of three methods : (i) Let
several holes be drilled, as at c, one large enough to take a
slotting tool, and slot all round with hand feed ; (2) Drill a hole
to take a vertical milling tool, shewn at c, and a few more holes
to save the cutter, and mill out the rest, traversing by hand, first
one side and then the other ; (3) best of all, is the same as the
last, with this exception : the cutter is held in a special form of
vertical mill, called a 'profiling7 machine. Here the bearing
carrying the vertical spindle may be made to traverse any par-
ticular curve by applying to it a copy of the same shape, and its
action is thus similar to that of the copying lathe. The curve
would thus be finished right off without further filing ; and the
ends may be taken out with a double corner tool, then finished
by hand. The die (Fig. 242) is ultimately fitted to the link by
careful filing and scraping, and both link and die (after broaching
the former) are case-hardened. (See Appendices I. and II., pp. 752

XXVI.    Governor Pullies (Fig. 258).— These are to be
machined as shewn upon the drawings.    The bosses are to be
bored by chucking in a dog-chuck, and the facing both of boss
and rim done at the same time.    Two settings are, of course,
necessary.    Next put the small pulley on a plain mandrel, and
the large pulley on an expanding mandrel, as shewn at page 155,
and turn the, rim surface in each case with parallel traverse :
then finish the curve with a hand tool.    The large pulley being in
halves must first be planed on its joint surface, as in the case of
the eccentric straps at Fig. 245, then drilled for the bolt-holes,
and bolted together for boring.   The keys are lastly t\ken out by

XXVII.   Governor Bracket (Figs. 259-60).— This is a
ratner more difficult example of lining out, but involves no new
principle, the only precaution of importance being very careful
levelling at every operation.   The casting is laid on its side, as in
the two views at B, and adjusted until tjie bush centres are of the

fn l^«





r v,


/*/ '


'    vii

f V       j f  .       ^   f I ,         ' 4 ^



' <,       I—r' -t


:'„,.,'      '  -  i      v      ,i

i * ,    t     .    - ;     ' f    J, -     >
; ,                 j t  > '   I         1     <

M   , **   *        '           J                .   3       '     V

? -1;:,,'»'        ?^         '>

w^'1^';/    i <      ,     t" ^ >   v
s , lf^« / i   4*   ! * i I ; *"   fi

5M>    i*    '•§   '

'/)        ' * f,    *»

<      I                      'iif

4 •, •      f j ' »    '    f
-     i' i     ,*

*^-;   t,^,K,/ 4*        I •>>*•      ^   t    • *    t    ,«   i'
^>r*-   ^^ '.'. -*"^*j 1s /   '^ /f»  1^1   4^ ,*   ti      i, ;', '*'

*i* ',.,4 "-.v*^, I it ««.f   ^.,'^f  tj  —,.   /*

*«,»»,*«,xf      It *  > ^-^ ^ t^.  >» , *. 2«,j.^^.
f^*«                          *• ^i f.#*1   *'ft«t^fr<i



Governor Details.


flat cheeks lined out These being shaped, the boss is next
marked oif and milled, and the hole drilled, the slot taken out as
in previous cases, finishing by hand. The key for the mitre
wheel is finally grooved with a milling cutter, The Sleeve B,
being cast solid, is first centred, and the thickness of the bosses
lined. It is drilled in the lathe, and then slipped over a
mandrel to turn, and to screw the end. The flat surfaces of the
bosses are next shaped across, and the space between taken out
with a tool of the exact width, The holes are marked off and
drilled, and the bosses filed round, after which the sleeve is fitted
to the arms c by chipping out the socket with cross-cut chisel
and finishing with a curved file much used by brass finishers,
called a 'riffler' (see Q, Fig. 262). Mark out key-way for the
weight E, and cut the same by hand.

The Nut for the sleeve is bored and screwed in a chuck, and
turned on a mandrel, and the octagon milled by fixing on a
dividing circle. Drill and tap for the side screw, but only file
out the corresponding slot in the sleeve after M is put into place,
and the nut advanced to give the requisite tightness, The
Lower Arm c is packed up as at H, Fig. 262, and the fork
bridged; then the centre and the flat cheeks are lined, the fork
centres struck, the lengths marked off, and the centres of the
bosses squared up. Next turn the shank, and slot or mill the
fork to the marked lines. Lay the arm in the position j, and
after scribing the centre line, strike the curves of the bosses and
pin-holes, and scribe the width of the fork. Shape and mill to
the lines, and drill the holes. The Radius Arm D is centred
and lined as at K, Fig. 262, and the shoulder line marked off, as
well as the commencement of the small curve to ball Set in the
lathe, and turn down the shank. Then prepare a template for
the ball, as shewn at L, Fig. 262. First turn to diameter as a
cylinder, and surface the end to length; then feed at 458, as at^
Fig. 262 j continue to halve these angular feeds until the ball
is approximately spherical, as tried with template, and finish with
a keen hand tool ground to the ball curvature. Mark the centre
of the ball while revolving in the lathe, and set on marking-
table to get the cross centre, as at M, Fig. 262. The boss is then
finished as usual, and the hole drilled through the ball The



>      1

256                           Governor Details.

Central Weight E is fastened to the table of a horizontal
boring machine, as shewn at N, Fig. 262, and bored with cutters
of correct radius. It is next put on a mandrel fitting the smaller
hole, and the outside turned to template. First the ends are
faced, then the diameter turned as a cylinder, and the rest is
obtained by various angular feeds, finishing by hand. The key-
way for fastening to sleeve B is to be slotted. The Bush F for
the weight, is to be bored in a chuck, and then turned on a
special mandrel, shewn at p, Fig. 262, being afterwards driven
tightly into the weight by means of a copper hammer. The
Nuts and Guard G are first bored, and afterwards turned on a
mandrel, being replaced in the chuck for screwing. The tooling
of the Lever H may be understood by reference to the regulator
lever No. I., and the studs j j are all examples of simple turning
and screwing. The Lifting Link K, and the Lifting Eye M,
need no special description. (See App. I. and //.,//. 753 and820.)
We now come to the Mitre Wheels L. For the machining
of these we may again refer to Fig. 262. Both wheels are made
of gun-metal and are exactly alike, boss included. After boring
truly they are placed on a mandrel, and the < blank' turned as at
A to a template which has been previously made with great care.
The teeth are then to be cut by means of a milling cutter. A
mandrel is provided which fits into the socket of the dividing
centre shewn at D, and the wheel set at such an angle that the
lower line of the tooth, ef, is horizontal. Looking in front of the
wheel, the work must be1 set so that one edge of the milling cutter
is in line with the centre. The radius <% of the cutter at c being
.made to fit the curve a of the larger end of the tooth, as shewn
at B, and the width b± of the bottom of the cutter made equal to b
at the narrow end of the teeth; a little consideration will shew
that the cutter will trim up one side of the tooth in such a way
that the smaller ends of the teeth d will be a little too wide at
this point, as shewn at* G. After all the spaces have been cut out
as at D and one side of the tooth, the work is traversed forward
and the other face cut as at B, after which the taper c of the teeth
is lined out as at G, using (i) a straight edge of tht forni shewn at
F, page 62; and (2) a template F, Fig. 262, These surfaces are then
dressed off with the file. (See App. Z, /£, ^pp- 753, 823,986.)

Steam Cylinder.                            257

XXIX, Steam Cylinder (Fig. 263).—The various opera-
tions are shewn at Fig. 264. The ends are first bridged, and the
centre found by reference to the outer curve of the cylinder flange.
Mark temporarily the height of the centre B. Adjust until the
top of the cylinder foot is fairly level, giving and taking with the
three centres at A and B. Scribe the horizontal centre line, B B,
all round, and square up the vertical line, c c; then strike the
circle D for boring. Line the heights of the steam and exhaust
flanges at E, and scribe the thickness of the foot at F ; line also
the thickness G of the bosses for the bolts. Scribe the height of
the valve-guide H, using a special piece of bent wire for the
scriber as shewn, and mark the heights, j, of the indicator

Set the cylinder upside down, as at M, and plane the foot.
Set upright, as at N, and plane the steam and exhaust flanges,
the indicator bosses, and the foot bosses. Now clamp between
angle plates on the planing machine as at p, and if these be
true vertically (as they should be) there will be no difficulty in
packing correctly; but if not, some care must be exercised, and
in any case the centre of the cylinder must be levelled longitudi-
nally. Scribe the steam chest face to correct distance from
cylinder centre, and plane with a front tool, p. At the same
setting the valve face may be planed at Q with long, strong side
tools, right and left, and the valve-guide also finished.

The cylinder must next be bored. This is done by packing
up, as at K, on a horizontal boring^iachine of the type described
on page-i6i, but of a smaller pattern. Bore right through, and
face the flanges by first measuring through the cylinder, as at j-.,
so as to leave an equal amount of seating on each flange; then
alter the tool to take out the bell mouth or larger diameter at
each end. Set the casting on end, as at o, and plane the stuffing-
box seatings so as to be level with the cover seating.

All the main faces are now machined, and the rest of the
lining may be done. At v the inner square is scribed on the
steam chest face, measuring from the horizontal, and vertical
centres, the latter being squared up with reference to tHe outer
edge of the flange. Take the material out by hand, or by slotting
tool, very probably the former. The steam chest coyer may be


f  -•


planed in the same manner as the cover at Fig. 237, and the
holes drilled, It may then be fitted to its place on trie cylinder,
and the holes scribed through, as at v, and centred. After the
front and back cylinder covers have been finished (next example),

/tfy ***•*                                    *!i *£**&

*•**                                                  r««tft^          I"    ^^

V^g,-;    ||                                   f«yj,    |t/ |     >»>*•*    4*^    |^f

tf»wfM*^fc   #t%i   4?                           V?/«Jf|    if*f                 **{*  %M*4f

4,4*.    t  '**. pi   pjtfy   *•+),             4?*   M,*»  i                 JHft^*11 f 4*u*

l**l, +>t 't*M



Slide Valves and Flywheel.

to proper* depth. Reverse the work, as at B, and set truly with
the turned face. Turn the inner surface and the smaller diameter.
Now set the cover, end up, on the marking table, as at c, until
the boss is horizontal. Having spanned the hole and found its
centre, scribe the centre line along the boss and on to the cover.
Scribe also the seating for the slide bars and the centres of the
gland studs. Now turn the cover through 90°, by measuring from
a square, as at D, and scribe the centre line across. Mark also
the slide bar seatings, and divide out the stud holes; then drill
the bolts and studs, and slot the slide bar seatings. Lastly, scribe
through the stud holes on to the cylinder flanges. The back cover
is turned, as at E and F, and similarly marked, and the gland
tooled as previously described at Fig. 238.

XXXI.    Main Slide Valve (Fig. 266).—Lay horizontal,
as at B, scribe centre of boss and thickness of valve, and plane
both sides.    Set up, as at c, till the bosses are level, and scribe
the centre all round ;  line also the top and bottom surfaces.
Turn to the position D till the boss is quire vertical, and scribe
the centre line round    From this, line the height of boss at top
and bottom..   Re-scribe the hole for the spindle at both ends,
which is much larger than the valve spindle, to allow for wear of
valve.    Next set up on an angle plate in the drilling machine,
as at E, till vertical, and drill the hole right through, knifing at
the same time.    Plane the top and bottom surfaces.    Line out
the ports by means of a template, and finish their edges by

XXXII.    Expansion Slide Valve (Fig, 267).—Set level,
as at B, and scribe the boss centres and the face.    Square up the
edge, measuring from the centre, and join this along the top.
Next set vertically, as at c, scribe the centre, and line thickness
of boss seatings.    Drill, knife, and plane as before, finishing the
edges to template.

XXXIII.    Flywheel (Fig, 268) requires very little descrip-
tion.    It is simply bolted centrally by the arms to a large face
plate, as shewn at A, the boss bored, and both boss and rirn faced.
It is next reversed, the other side faced, and the rim turned, as
in previous similar pases, the curved surface' being given by a
careful hand feed.    (For keyway cutting, see Appendix L>$* 754.)

^ J


Mr-r f





v                                                        ^




Bed Plate and Brass Work.

XXXIV.    Bed Plate (Fig. 269).—This is not too large for
a planing machine such as was described at page  169.    It is
better to plane the under edge, so that the bed may rest more
perfectly on the stone or brickwork.    The casting is therefore set
upside down on the machine, and the ends clamped till the two
side edges are planed; the clamping is then removed to the side,
and the end edges planed with a short stroke.    The bed being
now set right way up, and held by its lower rim all round, must
next have its seating marked, so as to plane off the calculated
allowance (the total depth is not of any consequence).    All the
seatings will be done at once, with a stroke the whole length of
the table.    The bolt and stud holes are to be marked off by the

XXXV.    Brass Work (Fig. 270) must be bright all over
the exterior, and have the interior bored at certain after-mentioned
places.    The Oilcup at B can be finished entirely by chuck
turning and drilling, polishing with the very finest emery cloth.
The Cylinder Cock, A, is cored throughout.    The main body
and the plug socket are both turned externally as far as possible,
but the central portion must be finished with file, and the corners
cleaned with  a riffler.    The socket and plug are respectively
bored and turned in the manner shewn at Fig. 254, the cock
then placed in the vice, and the plug ground to fit, with fine
emery powder and water, by rotating backward and forward with
a wrench upon the shank.    The screws are chased, and the flange
drilled; and the whole polished with fine emery cloth.   The union
nut, after finishing, is slipped over the copper pipe, and the conical
nipple then brazed to the latter (see page 86).

The Sight-feed Lubricator, D, is the only form now used
for slide valve and piston lubrication. The oil-chamber, #, is
fixed in any convenient position, and two connections made with
the steam pipe as shewn. Having filled a with oil, and the sight-
feed b with water, the valves c and d are opened, as well as the
two steam cocks, and steam being condensed in the coiled pipe,
forms water, which enters a afnd displaces the oil, forcing it lip
through the glass sight-feed chamber drop by drop, it being seen
rising through the water in b, than which it is specifically lighter.
Reaching the steam pipe, it is carried by the steam to the slide


!• I

1     'I'




valve and cylinder; e is a non-return valve, and f a drain cock.
The various parts are bored, screwed, and polished, and then put
together. The steam cocks are cast with a core, and are pro-
vided in casting with a small boss placed on the bend to assist the
centering in the lathe; this boss is shewn dotted.

The Indicator Plugs, c, are next turned and screwed.

Erecting.—We may now collect all the engine parts for the
purpose of erecting, as follows :—



L    Regulator Lever        ...        ...        ...    228

II.    Bracket for Regulator Lever ...        ...    229

III.    Pins and Washers for Regulator Lever    230

IV.    Links for Regulator Lever'......    232

V.    Regulator Valve Spindle       ......    233

VI.    Nut for Valve Spindle          ...        ,,,    234

VII.    Regulator Valve         .........    235

VIII.    Regulator Valve Box.........    236-7-8

IX.    Valve Spindles: Main and Expansion    239

X.    Expansion Eccentric Rod     ..,        ..     240

XL    Main Eccentric Rod ...        ...        . .    241

XII.    Intermediate Valve Rod       ...        ...    242

XIII.    Guide Bracket for Valve Rod         '...    243

XIV.    Eccentric Straps and Sheaves           ...    244-5
XV.    Slide Bars       ............    246

XVI.    Slide Bar Bracket and Distance Piece    247

XVII.    Crank Shaft Bearing......        ...    248-9

XVIII.    Slide Blocks............    250

XIX.    Gudgeon............    250

XX.    CrankShaft......        ...        ...    251

XXL    Connecting Rod        .........    252-3

XXII.    Cross-head      ......        ...        ...    254

XXIIL    Piston Rod     ............,    255

XXIV.    Piston ...        ......        ...        ...    256

XXV.    Radius Link   ...        ...        ...        ...    257

XXVL    Governor PuHies        ....       ...        ...    258

XXVIL    Governor Bracket      ...       ....        ...    259-60


Erecting the Engine.









Governor Details
Steam. Cylinder
Cylinder Covers

Main Slide Valve       ........

Expansion Slide Valve


Fly Wheel       ............    268

Bed Plate       ............    269

Brass Work     ...        ...        ...        ...    270

Bolts and Studs (not drawn).

The Erector is how to b'e provided with a * General Arrange-
ment/ or complete drawing,of the engine, in plan and elevation,
having certain principal dimensions supplied. This drawing is
given in Figs. 271 and 272.

The Bed of the engine is slung, and lifted by travelling crane
into position on blocks of wood, as at #, Fig. 273, and then
levelled with wood wedges and the aid of the square shewn in
Fig. 196; the cylinder and bearings then adjusted on their
seatings approximately; the back and front end of cylinder
bore being bridged with iron bars, the first having a small hole
drilled centrally arid horizontally, and the second having a
central notch in its "upper edge (see A and B) : a strong, fine
string b is knotted and passed through the hole, .and carried to
the front of t'he bed, where it is pulled tight and wrapped round
the support c'• the latter being set with one edge agreeing with
centre line of cylinder, as measured from the bearing seatings,
and having "notches, as at i), to hold the string at the correct
height. This' string 'constitutes the main centre line, and the
front of the cylinder is adjusted to.suit by tapping the casting
with a hamnier, then clamping firmly to avoid accidental move-
ment'           "

The Bearings are next adjusted Pass a long straightedge, %
through the brasses, and support it on level blocks till its upper
surface nearly'touches the string. Clamp the large $4uare F uport
E, near the string, and support on block G; then prerjare a lath B,
to measure ' the length from cylinder face to edg& of bearing
brasses, and* mark distances on the straightedge B, on each side

Erecting the Engine.                         269

of the square, up to the bearing faces (shewn by curved arrows).
Adjust bearings till (i) straightedge touches measuring lath;
(2) square touches string along its whole length; (3) face of
brasses is lineable with measure of straightedge j and (4) straight-
edge exactly touches bearing brass throughout its length. Then
mark the bearing stud holes through upon the bed, and do the
same with the cylinder holes.

The Slide Bar Bracket must be placed with reference to slide
bar length. It is wisest, therefore, to temporarily fasten the front
cylinder cover and the two bottom slide bars. When all are
together, as at dy the bracket is set to central position by squaring
up from its top surface to the string, and the stud holes traced

The Valve-Spindle Guide Bracket must also be true with
regard to the spindles, so these are put through the stuffing boxes
wid the bracket, and the latter adjusted by measurement from
cyliader face on the one hand, £*, and from the string on tfie
other hand, using blocked-up laths at h. If the spindles do not
slide truly, a slight readjustment can be made. Examine also for
appearance regarding seating, then scribe the stud holes. The
governor bracket comes to the Erector fitted up entirely with
governors, links, and pullies. Set up in approximate position,
and measure the distances j from the boss faces to the string,
these being the most important; adjust to these, and also to
distances from cylinder face (<?) and crank bearing (K)J, Then
test, by measurements at K and G, for parallelism of pulley
spindle, and mark off the holes.

The holding-down bolts are lastly marked, all the parts re-
moved, the circles centre-punched, and the Bed Plate either taken
to a racial drill, or drilled by ratchet brace, the former being
preferable. Tap all the stud holes and insert studs; then return
the bed to its erecting position, which need not this time be
level, the main adjustments having been made. And now the
various pieces are to be put in place in the order we sWl mention.
First the Cylinder is bolted down, and the Front Cover put on,
the Piston Inserted, and the Rod passed through; then the Slide
Bar Bracket, and the Bottom Slide Bars. The Bearings come
next, and when fixed have the Crank Shaft laid upon them, with


?  ;




... f i



' a



:   I

t               '            IT «i           I


272                        Setting the Eccentrics.

red ochre applied to its journals. Being turned round in the
bearings by means of the temporary handles, L, it is lifted away,
the brasses scraped, and the method repeated until a perfect fit is
obtained. The last time the shaft is removed, it is taken to the
marking table to line the eccentric key ways. The angles for each
sheave are shewn at M, being known after design, and are there
called a and ft; x and y are therefore found from a table of natural

For   oc = radius of shaft x sin a,  and y = radius of shaft x sin fi.


	Dcg.    Sine
	35 j "57357

	35* I -58070
	24 1
	36   -58778












	"5* 5°3

	43 | "68200

	' 35836




I 1*

The heights x and y above or below centre line have to be scribed
by laying the crank webs vertically or horizontally a$'~at p and w,
and the distance Q also measured, giving the centre line between
the two eccentric rods. Slot-drill the keyways. Then drive the
sheaves upon the shaft, to which they should fit tightly; put in
the keys, and replace the crank in its bearings, (It may be noted

Setting the Valves.

that the copper hammer should always be used in these operations.)
Bolt down the bearing caps.

Fix the Valve Spindle Guide, valve spindle Stuffing Boxes, and
Valves, also the Governor Bracket with gear complete. Twist the
valve spindles round until the valve screw is placed symmetrically
with regard to the valve; then measure for equal play either side
of the guide bracket. In the case of the expansion spindle, put
in the intermediate rod, and let the lifting link be vertical when
valve is at half stroke. We shall proceed to set the valves; so
to aid us in turning the crank to its various positions, the flywheel
is driven on to the shaft, and there keyed. The governor pulley
can be put on afterwards, being in halves.

It is convenient to find the position of the main slide valve by
the aid of a thin wedge of wood, R, which is tried in the port on
the horizontal centre line, and on removal measured. Put the
main slide to open to 'lead' at the front of the cylinder, the
amount being known; place the crank horizontal, as taken from
the seatings, and put the crank pin to the front, as at s. Now
measure with a lath the length from valve spindle pin to nearest
edge of eccentric sheave.

Set the valve for lead to back of cylinder, place the crank in
horizontal backward position x, and measure the length as before.
The two lengths obtained should differ only by a very small
amount, and, being averaged, the length ©f the main eccentric
rod can be found. During the preceding operations the expansion
valve can be slid to one side or the other for convenience.

The expansion slide must be set centrally. We first move the
main slide to opening position at front and back part alternately,
and each time measure the distances u on the valve spindle. By
setting the spindle ta half the sum of the two measurements at u,
the valve will be central ^The expansion valve is now moved till
midway between the main valve ports (v v), and its spindle
measured as at u. Set the crank webs upright, as at w, with
|jf                   straightedge and plumblines. Take the distance, z, to' eccentric-

centre, found thus:« >

z » radius of eccentric circle x sin /3;
and move tike expansion spindle back at u by Ms amount; then

;   .-     :   •   •   ,                 '                         .                '                                           *                                                                       XJ


Templates and Jigs.


measure length for eccentric rod between pin of radius link and
edge of sheave. Reversing the crank, as at x, the valve is moved
to the front by the same amount, z, the length again obtained,
and the two averaged. In our description of the machining of
these rods we supposed the length already given; but it is always
found for the smith in this way, though often the rods are finished
in two pieces, and afterwards welded to correct length.

Put the valve rods in place, also the crosshead, connecting
rod, gudgeon, and slide blocks \ connect up to crank pin, having
previously fitted the brasses to the pin by scraping, and bolt down
tfre top slide bar with distance pieces between. Fix the regulator
valve box (previously put together), the cylinder cocks and lubri-
cators, the steam chest cover, and the back cylinder cover, making
all joints with red-lead ' putty' between. The putty is a mixture
of red and white lead, softened with boiled linseed oil. After
covering the joint surface, a piece of soft hemp line is laid once
or twice round, and the cover then put on. Portland cement or
asbestos discs are also used.

The last stage of all is to carry away the parts to their per-
manent position, and bolt down the-whole to its stone bed;
connect up the steam and exhaust pipes, and get up steam.

We shall now conclude with one or two general points.
^ Templates and Jigs.—The former have been sufficiently
explained in Figs. 253, 264, and 266. They are used very
extensively in much repeated work, thus saving a great deal
of time humarking off, and they take a variety of shapes. Jigs
are an extension of the template principle. Instead of thin
plates, castings of an inch or so in thickness are used, supplied
with holes where needed, the object being to guide the ilrill to its
proper place on the work without the necessity of lining-out at
all An example of the application of this principle to a cylinder
cover is shewn at Y, Fig. 273. (Set Appendix //.,/. 820.)

Hobbing a Worm Wheel.—A cutter for forming spur*
wheel teeth was given at Fig. 186, and a method of cutting bwel
teeth at Fig, 262, Worm-wheel teeth can also.be cut by first
turning in the lathe a worm of the correct shape, and of good
steel. This is then fluted to form a milling cutter, and is termed
a fob in the workshop. The operation is then much the same as




i* 14'

I i




Dimensions and W. G.                       -277

that described at page 58. The spaces are first cut on .tire angle
with a spur-wheel cutter, and the finish given with-the hob by
placing both wheel and worm in position, as at Fig. .72, and
rotating the latter on a milling spindle. (See App. //.,/.* 823.)

Dimensions.—In most workshops the inch is divided into
vulgar fractions in the common way. But in dealing with work
of great accuracy, or where small differences are to be repre-
sented, the above divisions have to be carried beyond sixteenths,
and then become cumbersome. To avoid this difficulty, the
decimal system of division has been used for a considerable
period in a few shops, and has proved a great boon, being easily
learnt by any workman, and its advantages greatly valued. We
have spoken of high and low gauges for interchangeable work.
Where these are used, the drawings are supplied with what are
known as 'plus and minus' dimensions. Thus all shafts, pins,
&c., are figured "002" larger than the size required, an inch pin
being 1*002"; and holes are marked "005" larger than their
pins, an inch pin requiring a hole 1*007". There,is an under-
stood plus and minus allowance of "002" on both these dimen-
sions, so that if a large pin and small hole come together, there
will be a minimum clearance of *ooi", while a small pin in a
large hole will have a maximum clearance of -009". For driving
fit, the hole and shaft are figured the same, and the kind of fit
noted. (See Appendix //.,/. 825.)

It was long ago found advisable to fix the thickness of thin
plates and the diameter of small wires by reference to a table of
numbers, which received the name of the Birmingham Wire
Gauge or B. W. G., and where each number had a corresponding
dimension. This table was readjusted about the year 1885 and
considerably extended, under the name of the New Standard
Wire Gauge, and has been shewn diagrammatically in Fig. 274, the
horizontal scale representing a length of half an inch, while the
ordinates are referred to the numbers on the left hand. The
actual gauge is represented at K, Fig. 273, being a steel plate
provided with slots of the correct widths.

Split Pins.—Half-round wire split pins are made in fifteen

•different sixes,  the largest being •$/',  f", and Ty,  and  the

remainder numbered i to 12, corresponding with W. G.    The

J    I

'j*l fin                      aifi                        pit

4t||i iff                                                 Wti

#t|i             t*f |pif

tff4     '4*1              H 40f fl  pt»* ? J*| J^ l»ij            f * J*|

H»           »|                     t 4 1»*                «f|§ f pit tflft

M|i         i» 'ir*|         lit *f              »|        fo        i

f          *tf|J tt|                          W||             |i»

put         atfi                         i*t                 j*#

»ifi                    4*4                                           4tt»

f*t 4t||                                    |*

f«i |m P                                 lint

t^if M              *<ji               Jtt          t               1 1|

IT                                       *|f

dtp in                              nt tf                        t li|

f»it         J4                                   irii               *if*|


t?                          fitir                                            |r<

«          Ji||                 |W|   jiff

JO           f   4*1   if                  III ti|                  W*

fli^{ltJ»l                     1*1                                            i^l

•II If          *ii|,ff                                    |*i           it**| frt                    *^|f i*i

%%jttii«l ^t||             *|| 4 |t %r                  $^|          fut §

%f                                        't*»i{                                A»| %j|if|4 ik'Ji

111                       If*                   atfl ja

F         ill   *fi|f«   lilii             Hi   ||

«>||| |lf                           tM*'*t'V*, l»|                               %4l^ll<^%

V/Vlf                -^M ^f                         (%^ii^ui4

I4IJi»i^li|                  III                4t|   ^            JJ*   *  #%*   Vl t|^

*ftW{IIt*|           |f ^Iiifffill      'f III              §;

MJ   ffc4i*iJtlli*t   Ji.ftjll JU*Jt                      ,|«f   ^|*%          ^i|l|/**   i»^n(|*

nil            WiiiiAJfitffiw |ni| if itf |i$ tiitp.«t|i r n^^^ t^iti^         -4||


'HA                      »


Plate Material

Iron plates retain the fibrous quality imparted to the bar,
and are therefore much stronger in the direction of the fibre than
across it. Owing to the secretion of cinder and scale between
the layers during piling, the finished plate must be carefully
examined for faults—(i) by eye, (2) by slinging from the four
corners and tapping, when the dull, ashy portions may be de-
tected by the non-vibration of sand sprinkled over the surface.



. 2.75.

JtoJUburvq MM Jb

Very bad plates are rejected, and the others placed in the scale
according to quality, thus evolving the various degrees of ' best,'
'double best,7 and 'treble best;' terms, however, by no means
sufficiently definite. The Yorkshire irons are made with great care
and a large expenditure of fuel, being also very carefully selected,
Steel Plates and Bars are rolled similarly. The ingots,
obtained as at pp. 79 to 8% .are, after casting, usually broken up,
piled, and re-heated, though spine ^authorities complain that this
destroys the homogeneity far which steel plates are admired, and
prefer to roll direct from the Ingot. The slubs or ingots should

Brands and Plate Quality.                     281

be well squeezed in both directions when made into plate. Steel
plates are much more reliable now than when first introduced, it
being clearly recognised that a certain amount of strength must
be sacrificed to ductility. They are not, therefore, considerably
stronger than* iron, but much more homogeneous or even in
structure, the particles being so thoroughly re-arranged when in the
molten state. Iron plates, on the other hand, are very various in
quality, even over one plate,1 because of the processes employed
in obtaining fibre. A test strip, either for plate or bar, rarely gives
an exact determination of the 'whole, while* the contrary holds
with steel. Steel plates are termed c mild' because they have little
more carbon than wrought iron plates ; they have some 30 per
cent, more strength than the latter, with one and.a half to twice
the elongation. The price of iron being also greater, it is not
surprising that steel is the only material now used for plate work,
excepting where continuous flame action (as in locomotive fire-
boxes) renders Lowmoor iron or copper preferable. Copper,
though more expensive, and losing its strength somewhat while
hot, is an "exceedingly good conductor of heat, and deteriorates
less under the action of flame, lasting therefore longer than iron,
while being more efficient

Steel, then, in a mild, homogeneous form, is the material now
generally used for all plate and angle work. Both steel and iron
are received from the rolling mill under the following forms:—

Plates         ---------

Angle Bais         L
Tee Bars           T

Channel Bars    n

H (Aitch) Bars            H

Flat and Square Bars •••
Round Bars                 •

Brands, qualities, and sizes of plates.—The qualities
of iron have been mentioned at p. 76, * common' being used for
bridges and girders, and the remainder for boiler work. Mild
Steel occurs in four qualities, thus :—

1.  Ship and bridge quality.      ,

2.  Ordinary boiler quality.

3.  Soft boiler quality.

4.  Superior quality (to resist flaime).

I %Ji                                                            *

I*                                **                in* 94                      ^

#k                                                           |4|                                            py

ptfi             #**|                 **

| i*tt m            ^                        j«» ***               «A0 j»

I                              fv

I                                                                      *4                        pV

* Ji» ^fp

*f t


'«f^«.    **f|    ^f

^»| |M^^i»f                            j#*

9 t

fe >'






M!                     *ij|                      «M{| 4||                      iflfi J*»J

*l                         f                      ftff                                   lift


Itjfffi** /


1   ''

«|           at)                                %|M,|   |*i  14ti                  r **•*>*>



'*/'"*./ /'**'//



4**                                 *

t    ' 'I   *?

i     «^'»|                           -'V

5    ?             '4   ,,   > p   v ,   '.»f j  i'  I, 3 *   I                 4 .*»

* 'J»       4'   '         ,
;/*'f<   »V,        r " (

r,?   f   «"t  t


*       ? A     t»',

,i^   /I
> A    t      ^



Hand-riveting and Caulking.

The Plater requires three chisels—the flat chisel, Fig. 201,
the cross-cut, as at Fig. 200, and another with curved profile for
chipping the edges of manholes, &c. Hammers are of three
kinds;—the fitter's hammer, Fig. 199, the sledge hammer, and a
riveting hammer with long head and small panes for places where
the sledge or the portable riveter cannot be employed. A
Riveting Gang consists of three men and a boy ; the boy
brings the red-hot rivet, which the leader inserts, as at r>, Fig.
279 • another man holds up the dolly, as at A; while the third
man and leader give alternating blows until the cheese head E is
formed. The leader then applies the cupping tool or snap B,
while the striker gives two or three smart finishing blows with
the sledge c. Work should be designed for machine-riveting
wherever possible, as hand work can neither make the rivet com-
pletely fill the hole or compete in cost.*

Before riveting a seam, the plates, if punched or drilled
separately, are brought into alignment by the podger and bolted
in one or two places; then the drift at A, Fig. 280, may be
applied and forced through by a hammer to clear out the holes.
Though of undoubted advantage if used temperately, the drift
is now banished from the best shops, plates being injuriously dis-
tressed by it when the holes are very untrue. When a joint is
to be broken, the rivet-heads are chopped off by the set B, struck
with a sledge, and the punch c applied to drive out the rivet
1 Caulking is the process of making a boiler joint thoroughly
staunch by burring up the plate edges with a blunt chisel or
caulking tool. In Fig. 281, A is the section of a boiler joint,
where the edge of the outer plate is bevelled at an inclination of
i in 8. Striking the tool B with a hand hammer a burr is formed,
and the rivet heads treated similarly, as at a. Severe caulking
with sledge diminishes the grip of the rivet and frictional strength
of the joint To avoid this a Mlering tool c is often used, but
there is no objection to caulking if a large number of light blows
be given. A Pneumatic Caulker will be described later. Caulking
the rivets is not considered necessary If hydjr&ulic riveting be
properly applied. (See f. 322, also Appendices IL md IVn$p.
826 and 949)

* See diagrams by Mr. Tweddell, prepared for his paper before the North-
east Coast Institution of Engineers and Ship-builders, p. 321.

Machine Punches.


v Punched v. Drilled Holes.—<• Formerly the holes were
punched in a boiler plate before rolling the latter into cylindrical
form, and alignment then obtained by very forcible use of the
drift. The holes were marked by dipping the end of a short
piece of brass tubing into white paint and transferring to the
plate; the puncher could not therefore give great accuracy, and
the plate needed considerable stretching when a pair of holes

made c half moons/ Later the centre-pop replaced the white ring,
and a * centre' punch as at B, Fig. 282, was used in the machine,
so that the hole could be punched with accuracy. The machine
punches thus took the successive forms, A, B, c, and D. c was
introduced to avoid distress of plate by giving a gradual shear,
and t>, Kennedy's spiral punch, still better carried out the idea of
c, as proved by actual tests. The bolster is shewn at E, to
support the plate while punching; and the size of hole (larger
than the punch) may be found by construction at o, a triangle


Crystallisation produced by Punching.

being drawn with sides as i : 6.    Then if d be diameter of punch,
and / plate thickness, d± will be the size of hole in bolster, or


The material removed from the plate is known as the
c punching,' or ' burr/ and during the operation a certain portion
is compressed into the surrounding plate, thereby increasing its
density and causing i distress ; ' the clearance between punch and
bolster hole is to prevent this, which it does partially. The dis-
tressed area is said to be small, and the distressment relievable by
rimering, annealing, or both. Dr. Kirk's experiments in 1877 on
the fracture of punched plates, shewed the crystalline or weak
portion varying between the two limits at F, Fig. 282. All this
was removed by subsequent annealing, heating to redness, and
slowly cooling.

But the question was raised : if the plates require such treat-
ment after punching, and alignment be not then obtainable unless
punched after rolling (very difficult with machines as made), why
not drill them at once and avoid annealing ? There is no difficulty
in drilling after bending, and further, the holes may be made
through both thicknesses of plate at once, thus securing accuracy
of position. Drilling * in position ' is therefore the present-
day practice, and we are not aware of any workshop where
punching is performed except for very thin plates, or for
roughing out man-holes, &c. After drilling, the sharp edge is
taken off by a countersinking tool, or rosebit, to prevent cutting
action on the rivet, caused by expansion and contraction of the

Shearing causes the same harm to the plate as punching,
and the edges should always be planed afterwards.

D Cramps as at A, Fig. 219^, are required by boiler
makers for temporarily fastening plates together, or for providing
a hold when slinging.

Machine Tools, as explained in Chapter V., are daily
gaining ground, to increase the output, while securing greater
accuracy and cheaper production. As in the- Fitting shop, they
were at first driven entirely by belts from a line of shafting, but
the intermittent demand renders hydraulic power more advan-

Geared Punching and Shearing Machine.         289

tageous. Mr. Tweddell advocated the almost universal appli-
cation of hydraulics for plate work, and fully confirmed his advo-
cacy of the system, especially where the power had to1 be taken
about to various places in succession. In all shops Riveting
Machines and Flanging Presses are now actuated by water
pressure; so also may be Punching and Shearing Machines,
though more often driven by shafting; while Drilling, usually
performed by shaft power, has been successfully attacked by
electricity and water pressure; portable hydraulic drills, under
certain conditions, having proved both efficient and economical.

Punching and Shearing Machines.—It is customary
to combine both operations in one machine, as a plate is seldom
punched and sheared at the same time. Fig. 283 shews a good
example of this tool, as made by Mr. John Cochrane, of
Barrhead, capable of either punching, shearing, or angle cutting.
A shaft A has fast and loose pullies at B, and fly wheel at c for
overcoming variable resistance. The power passes, by pinion
and wheel, D and E, to a second motion shaft F, and in like
manner, by wheels G and H, to the main shaft j. The shaft j
has eccentric pins KK formed upon its ends to give a vertical
reciprocating motion to the slides L and M, the former carrying
the punch, and the latter the shearing knife. Dies upon the
pins KK prevent undue wear, and the fork N prevents the rising
of the plate when the punch is withdrawn. The shearing knife
always moves while the driving shafts revolve; but the punching
slide L is driven from pin K through the hollow die P and a
cam piece Q, the latter being connected to a handle R. When
R is upright the downward motion of P is transferred to L: but
if the handle be laid on its side, so also is the cam; P then
moves freely without pressing upon L, and no punching occurs.
Thus by changing position of R, the workman has ample time to
set his plate, while the shafts still revolve. The dies are hard
steel, and steel plates in slide M receive the wear. The angle-
shearing knife is fastened to a rocking lever s, actuated frotn
shaft j by an eccentric T, having ball and socket connection to
the lever* Here, again, the withdrawal of a sliding piece t; serves
to stop the motion of the knife, which is necessary with hars,
though not vrith plates.



Hydraulic Punching and Shearing' Machines.       291

At Fig. 284 is shewn an Hydraulic Punching and
Shearing Machine, designed by the late Mr. R. H. Tweddell, of
Westminster, for performing the same operations as the foregoing
by means of water pressure. In this example there is no reason
why the three parts should be combined except to save floor space.
A cylinder and ram are required for each operation : A for punch-                        I

ing, B for angle-cutting, and c for shearing; there are also the                        J

lifting pistons at I) D D. Water being supplied from the accumu-
lator pumps at a pressure of 1500 or 2000 Ibs. per sq. inch,
two pipes are connected with each cylinder, one for ' pressure'
and the other for ' exhaust,7 marked P and E respectively. The
valve boxes at F are supplied with piston valves (worked from
hand and foot levers j and K) to control the supply and exhaust;
but a constant pressure, on the pistons D D, causes the rams to
rise when water is exhausted from the main cylinder. A
small lever G, moved by ram c when at the end of its down                        f

stroke, is connected to a screwed rod H, having adjustable discs,
which restore the levers j nd K to the horizontal position,
stopping the water supply and the movement of ram c : this is
known as cut-off gear. Two overhanging cranes L, L? support
the plates while being operated on.

The Multiple Punching or Shearing Machine in
Fig. 284^, on Tweddell's system, has been designed to prepare
plates required in forming wrought-iron pipes for conveying
water or oil across country, and known as 'pipe lines;' it is also
useful for ships' funnels and masts, and for girder work generally.
A shearing blade or row of punches can be attached at will; the
latter being shewn in operation at A. The punches are set alter-
nately low and high, so that th^ punching resistance commences
gradually, and they are attached to a beam B capable of vertical
movement Downward motion is obtained by a ]bl!ward travel
of bar c, whose lower rollers press upon beam B, while the upper
ones re-act upon inclined planes D, D, b, fastened to the framing,
The working ram E (see enlarged section) moves bar c; water
entering the cylinder F from behind, and connection between
c and E being made with a toggle o, to allow for vertical travel,
K is the valve boic with piston-valve moved by lever j, and the
cut-off is effected automatically by the bell crank K, as



294                  Plate-edge Planing Mac/line.

previously described. A fixed ram N on the top of the
framing, has a cylinder M in the- form of a girder, to which a
constant water pressure is supplied, and the girder is connected
by bolts to the beam B, so that a rise of the latter takes place
whenever the main cylinder is opened to exhaust. The angle
bar p prevents the plate from lifting, and L is a stop valve.

A Plate-edge Planing Machine is shewn at Fig. 285,
having a long table A, upon which the plate is clamped by
screws B B. The tool c is fixed in a cylindrical box, provided
with handle r> resting on stops, so that direction of tool point may
be reversed at either end of cut, shewn by the arc E j this is
performed by the workman, who travels on a platform F attached
to the saddle v. The latter has a hand-wheel and screw G to
set the tool, while the wheel H, turned by hand, gives vertical
feed. The saddle is traversed by screw j, driven from the
countershaft K by gearing: while K is provided with fast pullies
M, N) and loose pullies L L L. When the forks are in the position
shewn, no work is done, but if the straps (crossed and open)
be moved to the right the saddle will travel to the left and
vice versd. Reversal is automatically effected by projections
p p striking the stops Q Q at either end of the stroke alternately,
thus moving the straps, decision being given by the weight R,
which causes a pressure between the rollers at s. The mid
position is fixed by stops T : and the standards are so arranged
at u that they overhang the work, thus allowing the admission of
any length of plate. One setting may serve for several plates.

A Band Sawing Machine is a very useful tool in a boiler
shop for cutting out plates of intricate shape, while straight plates,
too thick to be sheared or punched, are cut by a circular saw
^rhen necessary. As these are so well-known in their wood-
working capacity, diagrams have been thought unnecessary,
v Plate-Bending Rolls, in their most common form, are
shewn in Figs. 286 and 287, the rollers being supported hori-
zontally. These are the design of Mr. John Cochrane, of
Barrhead The lower rolls A A revolve in fixed bearings, while
those of the upper roll B are lifted or lowered by the screw c, the
worm wheel D acting as a nut, while the worm is turned by the
spoked wheel E, A A are the driving rolls, and the gearing ig very

" *

Plate £> eliding Rolls.                         297

powerful, consisting of wheels and pinions F G and H j, the last
being on 'the driving shaft, while M MN connect the rolls. The
pullies are driven by crossed and open straps, to obtain reversal,
K being the fast, and L L the loose pullies, so that either strap
may be put upon K alternately by a foot or hand lever attached
to the forks (not shewn). The plate to be bent is placed upon the
rolls A A? B lowered till a grip is obtained, and the machine set in
motion. When the plate has been drawn nearly through, the

machine is stopped, and the wheels EE given a slight, advance,
the rolls then reversed, and the plate brought back, and these
operations repeated until B is depressed enough to give the
necessary curvature. When the plate is' bent into an en tire "circle
It cannot be released at the front ; so the tof> of the standard is
made separate at P P, and the bolt Q turned down as shewn
dotted, when portion P F may be swung round horizontally upon
pin fc, leaving the bush g upon the roller B. The plate may then
be withdrawn. It should be noticed that the sides of the bushes
are curved in plan to radii from the pia R. ' •


Rolls for bending Section Bars.

Vertical rolls are often used for long, heavy plates, and are
said to be less expensive in operation, while giving truer finish to
the end of the bent plate. This last is the principal difficulty
with all rolls, the entering edge, to six inches deep, being always
set by bending while hot with wooden hammers. Except for this,
the plates are never heated for rolling, even up to i\ inches in
thickness, for in such cases the radius is proportionately larger.
The weight of plate is eliminated by the vertical method, with less
fear of obliquity of curvature. Long rolls are often slightly bellied
at the centre, to take up spring. For the heavier plates an hydraulic
bender, introduced by Mr. Tweddell, seems very likely to super-
sede rolls. It finishes the plates to a truer circle from end to end,
and there is no limit to plate thickness, or risk of fracture by too
rapid feed. Butt strips can also be bent truly to boiler curve.
The tool is similar in design to the multiple punch in Fig. 2840,
but the girders are placed vertically, and suitable dies inserted
instead of the row of punches.

Plate-straightening rolls are similar in construction, but there
are some four rollers at top, pressed down simultaneously by
connected screws, upon three rollers at bottom, and the plate is
passed through and through till truly plane.

Rolls for Section Bars (Fig. 288) have their axes vertical,
and are placed upon a table A, which is sometimes conveniently
set flush with the ground, with a pit for the gearing. They are
driven by the usual fast and loose pullies F and L with crossed
and open straps for reversal, These actuate a worm and worm
wheel, B and c, and a spur pinion D on the axis of c gears with
wheels G G on the roller shafts. Thus BE are the driving rollers,
and H the bending roller, with a screw j to bring its bearing closer
to the rollers E E, effected by turning the nut K. A ring or angle
bar is shewn bent to a circle with an outward flange—an inward-
flanged ring being obtained by turning all the rollers upside down,
and other sections by special rollers. Finally the ring is removed
and welded with a glut-piece.

V Flanging Presses.—It being always advisable to dimmish
the number of joints in a boiler, the end plate is usually flanged
or bent over at the edge to form a ledge for the shell-plate, while
stiffening itself considerably.



'Piedboeuf Flanging Press.

Plates were formerly flanged entirely by hand, being moulded
on cast-iron forms by blows from wooden hammers, as at Fig. 278.
This method was slow and expensive, and two kinds of hydraulic
presses are now used, (r) the 'PiedboeuP* press for flanging at
one heat, a very effective tool, but requiring separate dies for every
separate kind of work; (2) the universal, or three-ram flanging
machine, invented by Messrs. Tweddell, Platt, Fielding, & Boyd,
and capable of either progressive or single-heat flanging. We will
take these tools in order.

The 'Piedbceuf' Flanging Press, on TweddelPs system, is
shewn at Fig. 289. It consists of an hydraulic cylinder A con-
taining a ram B, which may be raised on the admission of water
pressure, thus lifting the table c, on which is placed the lower
die D. A girder E carries the upper die R, being supported by
guides F F, provided with nuts for the adjustment of E. The girder
G supports the central cylinder A, and four cylinders, H H, con-
taining the 'vice' rams j; and as it is necessary to move the
cylinders H H to varying distances from the centre, the pressure
(or exhaust) pipes are trained through three-quarters of a revolu-
tion between their connections at the pipe circuit K K and those
of the cylinder, so that the pipe is not strained materially when
the positions of H H are changed; in addition there are sheaths,L L
to prevent snapping at the unions. The valve-box M has two
hand levers; N for controlling the vice rams, and P for the
flanging ram. The two dies are shewn ready for flanging a tube
plate Q, which has been made red-hot and laid on the lower
die D. The vice rams are first advanced until the^pjate is held
against the upper die R ; then the flanging ram B is slowly raised
and the plate made to assume the dotted form. The levers being
reversed, the plate may be withdrawn. These, presses are made
large and powerful, but are not used for plates beyond 7 feet
diameter, and rarely up to that

Universal Flanging JPt£$3 (Twedddl's system). — This
very useful machine is sHewix at fj$g. 390. There are two vertical
rams, A acting as vice ram and known as the * elephant's foot,'
and B for flanging the plate on what is tooro^'aiB tl^e ^.^progressive
system.' A third and horizontal ram c gives trie finish, and a



4%       ^»

•T    j

*%    •<%

%    >






,f     :       '    «

.     .yr-TT-------        jt,

i j-v r .  i ^



Universal Flanging Press.                     303

fourth  ram   D raises and balances the vertical  rams A and B,

having a constant pressure supply; so that the rarns A and B only

rise when opened to exhaust, one or other, or both.    Yet a fifth

ram E serves as vice during single-heat flanging.    Referring to the

enlarged sections, the ram A is seen to be hollow, riding upon a

smaller fixed  ram  F.     Ordinarily  the  water   only  enters   the

annular space round the small ram, but on releasing plug G it

passes down the centre tube and then exerts a pressure on the

whole area of the large ram, a variable power being thus obtained

The horizontal ram c is of piston form with a tubular continuation

to a smaller piston H, upon which there is a constant pressure, so

the return is effected when c is opened to exhaust.    Any special

forms  of dies may be applied at j,  K, and L, and the guide

bracket M is removable.     The valve-box has five levers,  each

working both pressure and exhaust,  i for ram A, 2 for ram B, 3

for ram c, 4 for ram E, and 5 for an hydraulic crane to lift the

plates (see A, Plate XIV.).    A plate N is being flanged on the

progressive method.     It is slewed by crane, laid on a curved

hearth (B Plate, XIV.), and heated for a few feet along its edge,

then transferred to the block P and flanged as described, rams

A, B, c being applied in succession.    This is done foot by foot

until the heated portion is all flanged;   a new heat then taken,

and the work continued as before.    When flanging with complete

dies, the upper die is fastened to the rams A and B, as shewn at R,

and the lower die placed on the table.   The hot plate being laid                     jj

on the lower die, the vice ram E is first-raised and the upper                     I

rams then lowered; the flanging pressure is therefore the differ-

ence of that upon the lower and upper rams. Any kind of flanging
can be performed by this machine by using suitable dies,

Drilling Machines, for boiler work, vary greatly in their
construction. Except for the Radial machine they are all
designed to drill ' in position/ and their form depends on the
kind of work to be done. When possible they are made ex-
peditious by the use of more than one drilling head, a necessity
in view of the krge number of holes to be drilled.

Radial Drill—This has been already described at p. 167.
Opinions, differ regarding the best construction, but in almost any
form It is &a extrenaely useful tool for boilermakers. An inter-


Special Radial DrilL

esting example is shewn in Fig. 291, designed by Messrs. Geo.
Booth & Co. for performing a variety of operations. The circular
table A, provided with worm wheel B, may be revolved whenever
the worm shaft c is connected to the driving shaft D by belt; at
other times it is stationary. A bracket E? fixed upon the bed of
the machine, carries a tool F through the medium of the two
slides G and H, each provided with hand wheel and screw, thus
giving adjustment in both directions. When, therefore, a boiler end
plate is fastened to the table through temporary rivet holes, and
the worm gear connected up, the tool F serves to turn the outer
edge, the usual back gear being seen at K. The power further
passes through mitre wheels and vertical shaft within the pillar to
the spur wheels L M, and thence through shafts N and o to the
drill spindle, the feed motions being as previously described
The simple drilling done on this machine is the taking out of
tube holes in the manner shewn at B, Fig, 169; but large flue
holes are made by using the head p and three cutter bars Q Q
held by set screws with removable cutters, forming in fact a large
pin drill. In all cases a hole is first drilled in the plate to receive
the ' pin' and steady the cutter, and the radial arm R being long
may be fixed to the bracket s when doing heavy work. But the
most interesting feature to the student is the method by which
large oval holes may be formed, such as those required as man-
holes. A short vertical shaft T is connected to the driving shaft
N by gearing of 2 to i, the same ratio as that of the bevel gear at
u. At the lower end of T is an eccentric stud adjustable within
certain limits, and a rod v connects this with the saddle. The
shaft T making its revolution in the same time as the drill spindle,
an inspection of the diagram at w shews that the cutter will be
compelled, by the movement of the saddle, to mark out a true
ellipse instead of the circle it commenced with, which will be
understood by comparing the numbers; of course only one cutter
bar can be used. The tube j may be turned ,rpund within the
base x, for fine adjustment, by, the worm gear at Y, but the
position of the arm R is first roughly obtained by releasing the
bolts zz. The lifting is effected by tlje $crew a, driven from the
central vertical shaft by spur wheelsr at ^, reversed or put out of
gear at will by the htndle d mqvefl, horizpr|tally. This machine


V   "


Drilling in Position—

therefore can perform no fewer than four operations—flue-hole
cutting, oval manhole cutting, tube drilling or other single
drilling, and boiler-end turning.

Although the foregoing is a very useful, it is by no means a
usual tool. The rotating table is more often placed on a bed by
itself, constituting a vertical face lathe/

Drilling in Position.—The plates of a cylindrical boiler
being prepared and temporarily connected, the rivet holes are
drilled right through the several plate thicknesses. If stationary
machines are employed they must be supplied with a cradle or
bed on which to lay the boiler, so that the latter may be turned
round on its axis, and thus present all portions of its surface
at various times to the drill. Obviously there are two principal
ways in which the axis may be placed, vertically and horizontally,
the latter being used for large marine boilers, while the former is
advantageous when drilling locomotive or Lancashire boilers,
though it has also been employed for marine work.

Drills with Boiler Axis vertical.—Fig. 292 illustrates
this type of drill: and its individual application (the drilling of
rivet holes in the flanges of boiler flues), will first be described.
The machine is the design and patent of Messrs. Geo. Booth &
Co., and is very ingenious throughout The flue is bolted, with
its axis vertical and central, upon the circular table A, and a
handwheel B, being connected to the table by bevel gear c and
worm gear D, serves as a dividing plate, its revolutions being
counted to turn the flue through any fraction of its circumference
between each operation. The saddles E F ride upon vertical
standards G H, and contain horizontal slides j K, for adjustment to
various diameters. / is the driving cone, and power is taken from
horizontal shaft L by mitre gear to the vertical shafts M and N :
from these the various motions are obtained. Thus the spur gear
and mitre gear at o and p give motion to the horizontal spindles
Q R, and from thence by mitre gear to the vertical spindles s T,
which turn the drills u u and v v by spur gear. The vertical
movement of the saddles is given by hand or power. When by
power, a worm on shaft N gears with worm wheel w, which
actuates a second worm and wheel at x, connected with the screw
Y by mitre gear. The mitre wheel on Y rotates within a boss cast


1    1


— With Boiler Axis vertical.

on the saddle, and has a plain hole, the connection with Y being
by key only. There are two nuts z and a upon the saddles, and
the screws b and Y move simultaneously on account of their
union by horizontal shaft at d. When, therefore, the driving
shaft L is rotated in its proper direction, so also are the drills u u
and v v, and a downward feed given to the saddle, as described.
The raising or setting of the saddle involves hand gear, the
capstan t turning the screws through pinion and spur wheel, and
the mitre gear before mentioned: but although the spur wheel is
keyed to its shaft, the worm wheel x is not thus secured, and is
only in gear with the screw Y when damped to the wheel ff-while
the nut a is carried in a socket, and is adjustable by mitre gear
to alter the relative heights of the saddles. Horizontal adjust-
ment is made by turning the capstans gg, each of which moves a
pinion within a rack, and the bolts h h serve as adjustable stops.
The drills themselves are worthy of notice. The upper ones, u tr,
are of the twist shape, but have a conical shoulder at the top,
forming a countersinking bit. The lower drills v v are for counter-
sinking only, and their feed, upward or downward, is obtained by
hand wheels and screws jj. The saddles, somewhat loaded with
all this gear, are coupled to chains passing over pullies kk to
balance-weights behind. In drilling a flue fixed upon the rotating
table, the saddles are raised by hand to approximate height, and
advanced horizontally by the capstans gg; then the stops h h are
set. The strap fork is moved on, the countershaft and the drills
rotated, while the feed wheel at x is clamped in gear. The hole
being drilled to proper depth and countersunk, the feed is un-
clamped and the saddle raised to allow the bottom countersinking
to be done by hand feed // Withdrawing the tools v v, the
dividing wheel B is operated to turn the flue by the amount of-the
rivet pitch, and the next pair of holes drilled as before.

Shells of Locomotive boilers are drilled by machines similar in
general build to that just described. A longer bed is needed,
that the standards a and H niay be advanced or separated by
a tommy-bar applied to pinion and rack. An internal dog-chuck
on the faqe plate grips the shell, and the dividing gear remains
the same* The saddles are materially altered, being similar to
those of the radial drill, Fig, 291, excepting that vertical screws

§i4                ,»4fi                                      iH|t                 u^ff/j»f *?

ti»ft j**                                 *«ft                  ***         ^ft

I**         **|i *w         t                %

I it % §                        ^4 1       I**    f»f

Hi *f                                      j4jf   > &                        ^%|   «*|  fj^iit^n   *»

d*|                           jffl               &                                        «% i*****

*t<                            »    *JV                                                 V#    ff    ^f/.|f    "tyfr   4Wty

*!«                                                't             rfuf-MH >; w   > w *t if(,h

*»•                           '*i  ^*^^| «*/- ,*jp |nw 'M ?</c« j*/tv*>*

I                                         ***'*»$ jj^f^*?*^   jr'fji   4»'ty*  ||4       i\<nttt\'tvyt

|i»|                     r                             I ||f|» ff'J  x   i# *   *<|^ *>,t   i  t'/i^n

3*M*>i* "I   I* fl/v^vfc  j*"» 5 ^",** >»"»*   1*9  ^^^if

jt|l  ^| »& "|            tM**^1!*  *5*«^**S   rt«^|« |^'>p-cj %*   f^  ; ^ n. ^'|

"*^ }> ^^i*l | f4t ****|»  }*«!** '^ JH*4  (^; fiwtjr^^ i.f ,r^" » f

fr|                                ^%|  <*£  %r   ^t #i   f t>»ti^ ^  J,HP \f /  **^ ' fH

4»|                           **'»|   •»r»f,<r                            *$   f-**'fci£»5   ^*^  ^'4"r*   t'*.'   rv'^fi

f?|                    >^|

^4|*       I"

^IMiI >4 4f «t
,|ij                  <>»>   nr <  «|*ft|m   '%  »  *'  >   iM.t/i*     %''('   « ( -I  # '

f    Illr'l*!    J, J ^*| t    1*    |rf   ,'w       jf^ I        ,v    f     ^     -**'?(

|wl«fHJ|i ii  ||         »/|;"-|iv-   -tf     /   '^'ir     v     , '»    '.
fit                                          |^|    JU4ii* »*f  "i    if     \\  f^in,     'i       'Yt?

l*tlv 141


^  ,1

lf/«   ^t|r*


J*t4t*i  J*« 4^»{r
r if  .^^i^U

* t     * J




I £ * ^ 7

*      *        «?       m

*     ~       *       £       *       <*      J£

*4     ^      KX     -^                *     *<

^,     ^        V        fr.       ^

**   ^     «     «



>** v







-*rf                               i ^

%                                             iff!                                                     jw||

|               i/|                    «l*

J>f%?                           £                                |#*|                         ft |                     *f^*»^                        i|>

,      ',                            ^f^*^«      J«V4                                                             *i                              0   4     *   *   /ft

t*,**rff   |,»    ^j>    *^|      i  i/          f    |*                                            f^     f|«*^l^^|

?«>v* jKtM»J ft*  J^ffi           *«

W-*<^jft't    iff*       ft  *#/'•$

*'     'I          •**   i*|*^|*7 f;-|


i* )   «

(T   t • f   * * ;   t t   i   «    i f f'^

1 1    , « i r     'i   jf fi            rt   » »f



Stationary Hydraulic Riveter.                 313

of holes. Considerable economy results from the application of
this machine, which is very well designed in Plate XIV.

Summing up, the great desiderata in boiler drills are rapidity
of adjustment and withdrawal of tool, and where possible the
introduction of multiple drilling.

Hydraulic Riveting, Machines.—It is to the late Mr.
Ralph H. Tweddell that the honour of introducing hydraulic
riveting belongs. No other method is now used, excepting
pneumatic and electric contrivances, which are now being
more employed: but steam riveting is entirely obsolete. The
advantage of hydraulics for riveting is very great: it is a power
that can be conveyed to great distances without appreciable loss,
it can be stored till wanted, and the steady and known pressure
on the rivet-head, coupled with the increase due to absorption
of the momentum of the accumulator weight at the moment of
closing, is just the action most desired. {See Appendix I., p. 754.)

Large Fixed Riveter.—This machine, on TweddelPs
system, is shewn in Fig. 298. The standards A and B are securely
connected by two bolts at c, and well designed to resist the
stresses caused in closing. A supports the cylinders, while B
serves as ' dolly,' carrying the tail cup M, and presenting a
nearly flush top surface, for the purpose of getting into corners.
The riveting cylinder y, carrying the heading cup, rides upon
a fixed ram x, and within Y is placed the ram u, which advances
the annular plate-closing tool v. The auxiliary ram x, of piston
form, receives pressure on either face for advance or return :
and the tank D, placed 20 feet above the top of the machine,
supplies the cylinders x and u with low-pressure water. The
pipe E carries this water to cylinder x, and the branch pipe R
passes to u, the check-valves Q and s in each case preventing
return excepting through the exhaust pipe L. The latter com-
municates with each of the piston valves, p, o, N, as does the
pressure pipe j; P being connected to the back end of the
cylinder x, through the pipe a: o with the cylinder u through
pipe A : and N with the cylinder x: while b is a constant pressure
pipe connecting j and the front end of x. K is a stop valve, and
z an overflow pipe.

We can now understand the action of the machine.    The


Portable Hydraulic Riveter.                     315

boiler seam being placed between w and M, the rivet heated
and put in from the side M ; lever H opens valve p to pressure,
and a right-hand advance is given to the ram x, due to the
difference of area of its faces. This pressure, assisted by the
head of water passing from the tank, through the check-valves
Q and s, carries forward parts u and Y. When w and v reach
the rivet and plate respectively, lever G admits pressure water at
o through pipe A, to advance the ram u, thus pressing the
plates firmly together between tools v and M. And now valve
NT is opened by lever F, and pressure given to T in turn, thus
bringing forward the cylinder Y and the cupping tool w to
close the rivet, the pressure obtained being due to the difference
of areas of the rams u and x, part of the water from u passing
into T through pipe j. The pressure should be kept on the rivet
until it cools somewhat, to secure a tight joint, and the three
levers are then moved to exhaust, when the pressure b pushes
back ram x, bringing u and Y to normal position, and lifting the
water up L into the tank.

Fig. 298 shews all the later improvements introduced : the
plate closing (in 1880) and the use of low pressure water to
fill the cyliriders (in 1890). The latter is very interesting, and
greatly economises high pressure water, which is only used as a
film on the back of the tank water, as it were, the fluid being
practically Incompressible. The plate-closing apparatus prevents
* collars' being formed on the rivet between the plates. In a
loo-ton riveter, 60 tons are applied for cupping, while the
remaining 40 tons hold the plates together, but ultimately the
whole 100 tons is applied to the rivet-head and plates.

Portable Hydraulic Riveters.—Although Mr. Tweddell
introduced hydraulic riveting in 1865, his invention of the port-
able machine did not occur till 1871, from which date Messrs.
Fielding and Platt, who then took up its manufacture, were
associated with him in the design of nearly all his later
hydraulic machine tools. There are two forms of the portable
machine known as the * Direct Acting' and ' Lever' types re-
spectively; their present construction being shewn ;n Figs. 299
and 300. Referring to the former, frame A is a rigid casting,
supporting a cylinder B with direct-acting ram c. There are three



Portable Hydraulic Riveter*

diameters on the ram; c and v to obtain two powers, while w acts
simply as guide for the cupping tool F. When the smaller power is
required, water pressure is admitted to the annular area D, but if
plug E be unscrewed it acts also on the back of c, the pressure then
being due to both areas c and D. K is the valve box, containing
the piston valve Q, capable, by means of the passages within it,
of connecting the annular chambers N and M, or of opening M to
L, where the pressure-water enters. G is the returning ram, upon
which a constant pressure is exerted through pipe H, and space N
communicates with the exhaust pipe j. The handle P acts on the
valve lever o, so that if the latter be moved to the left, space M is
uncovered and pressure-water enters cylinder B ; but if o be moved
to the right, spaces N and M are connected, and the cylinder
water passes out to the exhaust pipe. The machine is slung by
chains R R from a pulley T, provided with worm gear; by turning
which from the hanging chain T, the frame may be set at various
angles to the vertical within the plane of the paper. Studs also
are fixed on the frame at the centre of gravity of the whole, on
which are placed the slinging pieces x x, and the worm gear at
s turns the frame in a plane at right angles to the previous move-
ment : universal adjustment being obtained by the combination
of the two motions. The space between the cupping tools may
be adjusted by the insertion of longer or shorter dies, or by pack-
ing collars; and the method of riveting needs , no further

Taking now the lever machine at Fig. 300; A and B are the
levers, the first carrying the piston E and the second the cylinder
D, while both are connected by the pin or fulcrum c. To avoid
another joint the curved cylinder was devised by Mr. Fielding, as
well as special tools for its perfect machining: two enlarged
sections of it are shewn. The pressure pipe is coupled at j,
where a sheath attached to the union preserves the pipe from
injury by sudden bending, and the movements of the machine
are not interfered with, for the water passes through a swivel joint
at K, through the coiled pipe M and the swivel N, then through
the pin at N and the arm Q to the fulcrum pin; another swivel E
and a short pipe x connecting c with the valve-box, u is the
exhaust pipe, led away as required, and tbe piston valve H is



alp jt»                                            imp p

ili        m                  up pm         mi           if

*.          ;«£§ ;ti|  P

|   pit                          ,Jiif

M|I         it it              §tp ^p         MI

^^p          1^1

H |ti#» ^ *| if ,,

*l» ^i§ f*-                        |Mit         ttfi ^

aft fc  pa** 4                             ^si            ^1 |Wi

i^            ^ii         ;t         ^i p* nit m        tt

»ll H;)                                                             |*|



I      ^i



J*/   If   I*  I'.

*!'*"'   ^|l  »^*%-»

]J     {  '    >  J  f

',    I       !'     I'M #

i,    ,.,';';'{«**              j   > (

r     i           i;    r/   s 1         i    f /,   '

Marine Boiler Shop.                         319

required is very apparent, the importance of ample provision for
lifting being a point upon which Mr. Tweddell always insisted.

Plate XVI. represents the interior of a Marine Boiler Shop.
B is a Stationary Riveter, exactly as in Fig. 298, and a circular
pit c admits a large marine boiler when riveting. As it is difficult
to obtain nicety of vertical movement in the travelling crane D,
an intermediate cylinder or Hydraulic Adjuster, E, forms a very
useful adjunct. The Progressive Flanging Machine F was shewn
at Fig. 290, and the crane A lifts the plate to or from the fire. A
plan view of the latter is given at G, where the dotted lines shew
the plate being heated. H is the Locomotive type of Marine
Boiler, much used for the smaller boats, the riveting of which is
performed as in Plate XV. A Marine Boiler is given at j, having
the furnace mouth riveted round with a small bear K, which also
joins the ' Adamson' flues at L. At M the boiler is being closed
by a powerful bear-type machine, having plate-gripping tool, and
hung from the travelling crane through the medium of1 the
adjuster N. The last-finished flange is here turned outward, as
^advocated by Mr. Tweddell, to secure good machine riveting
throughout; but as many makers prefer an internal flange, to
save cargo space or reduce weight, the riveter at p has been
recently devised. Jt is slung from its centre of gravity, and the
free arm lowered into the boiler, as shewn dotted. When raised,
the latter serves as * dolly/ and can be adjusted in length to suit
various diameters of boilers. A hole must be left at Q, to be
covered afterwards by the plate carrying the central nest of tubes,
the final riveting of which is performed by hand.

The diagram in Fig. 301 shews the arrangement of hydraulic
tools on TweddelFs system applied to Shipbuilding. A is a keel
riveter, supported by parallel motion and balance weight, so that
it may be raised or lowered to reach the keel in any position, yet
remain with jaws vertical The gunwale riveter at E is similar in con-
struction, H shews a small travelling jib crane, carrying a direct-
acting machine for riveting the combings of hatchways, j and K
are Hydraulic winches, and G a punching or shearing machine. B *
is another jib traveller carrying the Large lever riveter c for finishing
the double bottom, and the machine B, supported by a crane with
two movements, is for riveting the keelson. A special carriage F

4        / rfl/              «^

«*           \|*4k   «#**«**

*»<£                                     ^»|                                              »Kyf   «'|

*    * if ^|

^^  *»vK(fj *                              *p l'i^«

#$<* -^                            *^i                    *^^v*


",   fc*»*f

fH ^ ft« *   4«          I

, i,,,....., ...*>j


f      11











1   100 0   180
 ft^   140 fc   I0°
 §      40 20

	^ y










55 45 55
 2o IS
















C     t>   £    f    G    H     /JHt-ABCOf:    F\G    H\ /








FRAMES                                              BEAMS
1 !° V     8
 §     4 S>      3


	s .










Pneumatic Caulker.

v Pneumatic Caulker.—-This, an American invention, was
first introduced in 1890. As shewn in Fig. 302, it was formerly
made by Messrs. Crossley Bros., and would do the work of three
or four men. E is a jacket held in position by the cylinder j,
screwed into the nose-piece F. The caulking chisel G is loose,
but placed within F when required. The piston contains a piston-
valve P, vibrating at right angles to the piston's axis, the slide
hole being closed by slips o o, dovetailed into K. The starting
valve R, when in the position shewn, allows the compressed air,
after entering at L through a strong indiarubber tube, to pass
through the piston by T and u, then harmlessly out by the
passages v and w; but if R be pressed down the passages v w are
closed and the machine operates in the manner to be described.
Key x allows the piston to slide,vertically, but prevents axial
rotation. Y is a passage from T to the piston, and T and u being
formed by flats in s, are not in communication with each other.
There are two passages from the piston to u, seen in plan at z zl9
while in the piston itself one passage j communicates with the top
of the cylinder and another h with the bottom. In addition, two
holes d d^ are made in the slips o o, and grooves e <?b /"/i af e in
connection with these holes at certain times. One other point
must be noticed—the hole g is the exhaust outlet when in
working order, but M fits the hole in the nose-.piece so that air
cannot escape when the piston is at the bottom of its stroke.
If, however, K be lifted to the top position, M will be found just
of a length to disclose an annular space round the curvature N,
and the air is free to pass out at g.

Having noted all the parts, we can now describe the working
of the tool. The workman, after placing the chisel G in the nose-
piece, holds the former with his left hand against the seam of the
boiler as at H, while with his right hand he grasps the boss st
pressing the head R upon it, thus practically closing the passage
u. The air passes through T and Y, but cannot get further.
Hole 4 is now in communication with passage t, however, so the
air enters the valve chamber from the right and moves P to the
left This allows the pressure to act through h on the bottom of
the piston, and the up stroke is made. While this air exhausts at
g, the hole d, being now in communication with f, the valve is

• t A v r / «   '. r    /   14  ;, i    i* {   ,    " t      -4*          :»,;•*,    « , i   ^ >    n

'<,<«},,*>,        "     f             r*    *„    ,•'        A    '          i   '       f1          ,'"',/

-,.   ff    .   1 {       ^    i t». i  •  ,, M      ft      .      ;!"•»•

'(f  V > ^      /      ft ,       t  ,  i      } JM ,4         4   V  ' $           |          "   $ f   /    > A   I       t              ,?    y            / jr f        1   "           f

,    v         , f«^   "f  "   * • &    8 i      , >    *       *  •«         * *     i i     ^    ft**'

'     T       J «*           >,    '      4    '           ^    «  ' / J     / f   '            > * t '       f             t     t *   M

'    '     *   ,    i* 4            4         f    I I '            f  < ,   /              .       t: f * I ,    , '   , I

f                  i, «   »"           /    «   t » < j       *          M             •,%<;<       •>'"«.           • s* *

* 1


Tube Beader, Cutter> &c.

Yarrow recommends the roller shewn at #, and advises that the
hole in the tube plate should be rimered to the same taper as
the mandrel F.

1 Tube Beader (Fig. 305).—H is the body of the tool, con-
taining three jaws at j capable of sliding radially, and moved out-
ward by the taper part of bolt K when nut L is tightened up.
This is done when the beader is put in place, the disc M serving as
steadiment. The collar N holds three rollers, placed at such an
angle as to do the work efficiently, and a ratchet wheel o is keyed
to N. P is the feed nut, and the ratchet arm Q rides loosely on N,
the latter being driven by Q, like the drill in Fig. 215. But there
is one depression b in the rim of the feed nut P, so that when Q
has, by its vibrations, brought N round by one revolution, the feed
nut is automatically advanced by a small amount. The firebox
ends of the tubes being excessively strained by the great variations
in temperature there occurring, beading protects the joint, while
the ferrule R, in addition, secures the rigidity not obtainable by
simple expanding.

* Tube, Gutter.—As it is impossible to gauge the length of
the tubes accurately beforehand, the tool at Fig. 306 becomes
necessary. Three bearings s s s, capable of radial sliding, support
hard steel discs x x x, which are the cutters. The tapered bolt u
advances these bearings outwardly when tightened up by the
nut v; this may be termed the feed. The tool body w has a
square at x and an adjustable gauge at z, b.y which tjie cutters
are set. The gauge being fixed, the tool inserted, and nut v
screwed up, a spanner on x rotafes the whole. Then v is
tightened, the operation repeated, and so on till the tube is cut

4 Ferrule Extractor.—As tubes have to be withdrawn and
replaced, and .the ferrule is the most troublesome portion to
remove, the extractor at Fig. 307 has been contrived to meet this
difficulty. The washer b is first placed against the tube plate;
then the set screw d released to allow the jaws ef to enter.
When all are in position the screw d is advanced to press the
jaws against the tube, and the nut g then tightened with a long
spanner ^nd the ferrule drawn out All the four foregoing tools
are supplied by Messrs. Selig, Sonnenthal & Co. *



Electric Welding.                       '     327

7 Electric Welding.—This important process, first intro-
duced in 1885, has proved of great advantage in satisfactorily
uniting pieces unattachable by ordinary means. Among these
articles are boiler plates, which must be our apology for intro-
ducing the subject here. Wrought Iron, or in a less degree Mild
Steel, were the only materials previously weldable, and even then
the joint had but 70 per cent, of the strength of the solid material
—a serious matter with crane chains, where every link is welded.
Scale might form between the weld, the heating could not be
seen openly, and might neither be even nor thorough; objections
all absent in electric welding.

Electric energy consists of two factors—electromotive force
(or pressure) multiplied by the current (volts x ampbres). If
this energy pass through a good conductor, nothing is observable
in the latter; if a bad conductor be presented, the current will
not pass ; but an indifferent conductor will allow some of the
energy to pass, while the rest is converted into heat on account
of the resistance, the amount of heat energy produced being
equivalent to the electric energy destroyed. The metals we most
desire to weld are in the class of semi-conductors, and there is no
difficulty in raising their temperature to welding point by the
electric arc; but the heating effect of a current is independent
of the pressure or potential, depending only on the value of the
current, and it follows that the energy from the dynamo must be
transformed, so as to obtain a low voltage with a high amperage.
Every one knows the galvanic battery and induction coil, where a
current of low potential becomes one of high potential after
passing the coil, though at £ sacrifice of current value, the energy
remaining the same. Transformers serve the same purpose,
being similarly designed, and it depends which *side of the trans-
former we are on as to what amperage we obtain.

There are two processes employed in electric welding, the
' Thomson' and the (Benardos/ named after Professor Elihu
Thomson and M. Von Benardos respectively. The first con-
sists in using the pieces to be united as the poles, and the second
in using one of the pieces as the negative pole, while the positive
pole is supplied by a rod of carbon, held, in/ the hand in the
manner of a soldering bit. The electric energy is obtainable in


Electric Welding Processes.

either case by one of two methods — (i) from an 'alternating'
dynamo, the ' current ' being increased by passing through a
transformer; (2) from storage or secondary batteries, which take
their energy from continuous dynamos. The welding apparatus
is not thereby altered. A general diagram in Fig. 308 shews the
direct method combined with the Thomson process, where A
is the dynamo, B the transformer, and c the welding apparatus.
Two wires are clamped in position at D, and end pressure put on
by the screws, the current switched on at E and regulated at F.
The ends of the wires are previously brightened, and a flux of
powdered borax interposed. After welding, the bar or wire is
removed and hammered to size.

Energy remaining the same, the following examples will
shew the variation in ratio of potential and current for various
purposes : —

1.  For arc lighting :              2500  volts at

2.  For incandescent lighting : 100  volts at

3.  For welding:

4.  For welding :

10 amperes.
250 amperes.
volt at 50,000 amperes.
volt at 100,000 amperes.*

No. 3 would weld steel bars ij inches in diameter in less than
two minutes, while No. 4 would do the same in one minute, ab-
sorbing 35 H.P., but only for a short time. The great advantage
of electric welding lies in the local character of the heating, which
prevents the spoiling of a finished piece of work.

We will now turn to the Benardos process, shewn in Fig. 309.
It is there worked by accumulators — the method most preferred.
The batteries being charged from a shunt-wound dynamo, they
are connected to a switchboard A, so arranged as to throw them
out in sets of five. From this board the current passes through
resistance coils for 'farther regulation, and then through the
welding tool B; the pieces to be welded, and back to the accumu-
lators. Fifty cells are usually employed, and, if two boiler plates
of about -j^- inch thick are to be united, the tool carries a very

* NOTE.— Only strictly correct in the Thomson process, where energy
absorbed is due to true resistance. The Benardos process uses the we, and
energy is required to produce light, vir., to volatilise the carbon and render it
incandescent : amovmting roughly to 30 volts in addition.

I ::





Corrugated Flues.


I '

through to the end k^ thence by brickwork flues, along the bottom
of the boiler to the front, again to the back end by the brick side
flues, and away to the chimney. The internal flues are therefore
at a greater heat than the rest of the boiler; this, producing
expansion, necessitates the introduction of elastic portions. The
flues, moreover, are in danger from collapse, for a cylinder,
although strong when pressed from within, is unstable when
pressed from without; so strengthening rings are applied at various
distances along the circumference. But as joints have to be
formed, on account of the great length of the flues, it is customary
to make provision for elasticity lengthwise, and rigidity of cross
section also, at these places, the most usual method being by the
introduction of the Adamson flanged seam at e. This joint .has
the advantage over other methods, of shielding the rivet heads from
flame, and a slightly projecting annular strip is placed between the
flanges for caulking purposes. The space between the tubes
being small, the seams are made to ' break joint' longitudinally,
so as to be easily got at when necessary. Conical ' Galloway'
water tubes are sometimes inserted, as at D, for intercepting the
heat more satisfactorily, the smaller end being passed in at the
larger hole. The flues are joined to the end plates by angle
rings, and their diameters decrease at k^ the connection being
formed by the conical portion /. The manhole edge at f is
strengthened by a riveted ring, always added when a large hole
is removed; and the mudhole n is similarly treated, a portion of
plate being left all round, on which to place the internal door.
Holes are cut for various fittings, as at #, g, and h. The circular
seams are single riveted, but double riveting is used for the
longitudinal joints, because any boiler receives but half the stress
longitudinally that it does in a circumferential direction (seep. 398).

Fox's corrugated flues, shewn In section at E, are extensively
used for the furnaces of many boilers, taking the place of the two
pieces //; while f is equivalent to the portion k. The corruga-
tions give not only strength and elasticity, but a larger heating

The proportion of length to breadth in the boiler shewn is the
largest allowed; more often the length is about two-thirds of that
(See Appendix 11^ p. 827.)

The Marine Boiler.


The Marine Boiler, as at present constructed, .is shewn
by the two views in Fig. 311. The number of furnaces depends
on the size of the boiler; in this, a large example, there are four*



The Marine Boiler.

A A. The combustion chambers are also variously divided, there
being from one to four per boiler; two are shewn at G G, each
having a plate c, to split and assist the draught. The heated
gases, rising from the fire at A, pass through the combustion
chamber G and tubes D, to the uptake, which is placed at E to
cover the tubes. The boiler is cylindrical, but with large flat
ends which require a good deal of stiffening, for flat portions in
all boilers are weak. There are two belts of shell plates i£ inch
thick, the first H and the second j, each being, on account of its
large circumference, divided into three, and connected by double
butt-straps, with treble riveting, as in plan at K« The division is
uniform and is seen on the front elevation; where F F F represent
the joints of the first, and B BB those of the second plate. The
circumferential seams are double riveted, as at L L, and the man-
hole is placed at M, with strengthening ring. Sometimes a
separate dome is connected to the top of the second plate, but
just as often the steam valve is applied direct to the boiler; in any
case the dome is simply a horizontal cylinder with dished ends.
The front and back-plates are divided into three, N, o, and P shew-
ing the parts of the front-plate, while Q, R, s are those of the
back-plate. N, P, Q and s are each f- inch thick, but R is only f inch,
and o is \\ inch. They are all flanged and riveted as shewn, o
being cut out a suitable shape to take the nests of tubes, while R
is rectangular. Where three plates overlap, the middle thickness
is drawn out as shewn at sz, which is a plan of the joints XT.
Longitudinal stays, for the steam space, are supplied by bolts
u XT, having large washers to distribute the pressure The plate o
is necessarily* stiffened by double thickness at the seams, but there
are also stiffening plates vv riveted on the inside, and stay tubes
w w, shewn by their nuts, support both plate o and combustion
chamber tube-plate x. The other tubes, ferraled at the firebox
end, and expanded at the uptake end, act also as stays. The plate
i? is stayed by bolts at Y Yl9 and the manholes are stiffened by
riveted plates at z. The three bolts marked YX pa$s right through
to the back-plate s, which is further strengthened, together with R,
by' screwed stays at a, which are bolts screwed their whole length
and fitting into holes tapped in the plates, The
chamber back-plate J inch thick, shown at G, is a' simple

The Locomotive Boiler.                        335

plate; but the tube-plate x, \\ inch thick, is throated to fit the
furnace flue. The top and side plates, J- inch thick, are in three
pieces, with joints at bbb, and wherever three thicknesses super-
pose, the mid plate is feathered, as at b. Screwed stays i£ inch
diameter, 7 inches apart, are fixed between the chambers at e and
at the sides, while the roof is supported by girder stays which
each consist of two plates resting by their ends on the roof seam.
Between these plates are passed collar bolts, which, after being
screwed into the roof and fastened by nuts, are tightened against
special washers on the girder. The furnace flues are of the Fox
pattern, flanged to the throat plate as shewn.

The Locomotive Boiler (Fig. 312) was the earliest form
of multitubular boiler, and has served as pattern for many other
steam generators. The firebox A is cubical and of -|" copper-
plate, thickened at: the tubes to ££". The back plate D is1 flanged,
and dished round j the firebox hole to the form shewn, the tube
plate c being also flanged. The top and sides are in one piece E,
and all these plates, being flat and weak, are supported from the
outer shell by screwed stays riveted over. The latter are £"
diam. and 4" pitch, and must be of copper, to avoid corrosion
by galvanic action, which frequently occurs next the firebox plate.
The shell top and sides are in one plate H, cut out as shewn at
HX ; the throat plate F is flanged to join the barrel and the firebox
•shell; and the back plate G is also flanged. The foundation ring a
serves as a distance or closing piece when fastening the shell to
the box, and a similar piece is required at z, called the firehole
ring. Mudhole bosses b b are welded on the solid plate, and
tapped for tapered screw plugs. A hole is cut in the top of the
shell at v for a double safety valve, and the plate stiffened by a
wrought-iron valve seating, From angles on the shell roof at w
are hung the sling stays x x, supporting the girder stays Y, the
latter being solid forgings, and the stay bolts taking the form of
tap bolts. T is a stiffening angle for the shell back, and p p are
expansion brackets which rest on the engine frame. The firebox
tube plate, besides thef ordinary screwed stays, has four palm stays
at s s, which are seen in detail at st. Two plates, K and j, form
the boiler barrel, and each makes a complete circle, the joints
being shewn in plan, well out of the water space. The dome L


'.'v,Vo ooooooo o o
'f O O O 00 O O O O O O O O O

i ooooooooooooo



|*l          %if|f

ff(        I I*                           ^*|                  |f   $*,»M*»»M*»         ****!

J|«   f    |*
|#«               |*                                    t|i          f|^l«|   J4f| |«

r            III   5**, |                    *i|i f#jf I"*?



fj|t^s»|                     |i»«w|4|

f                                             JlWO»J    M    ><**,!

n   «  i*^ jty»**iv   •*'!>           *«

>f|? |              -J4f |         II    |t   t?   *ȣ''* jM'l

I*    4H»'»*||i*» f     '"|S     ***|     4 ^f^t-Al *     Mf"*

i n   j/t/tjJH'^i I*      I »f nj**6>#

I*               'I              fJ,,%|!J^    1*^    fc*^4     ^**^|

I*'0<!M}»   ^<4F.P %^ll"i/

f        fill   |i^**^|if j*t*t^fi  v  u"   t       |#,

I   1*1                   r   '-i

III     Pi |   n*1                J^| |

|jj  tit i           r  p^s  r***'    '#f    ^

««ir ti'> »  *i%
*^^i « **^'
fi           i|'|rw

j?|l  ni   i.
.jfj^        ,         tfi

*^jfiiri|            j*|i f^f

M                 up           '•***«*  |»»i I

»,tf$f^  fr^tjiaa^***^"1    *
*i         «>f i   *«i |        ^i* i C


'« •;".*'*  ", V* If    ',*

'ir i* ** *i *;t!   ; .j |;   »*•  >    4   *|« j'^tt* i  4 ^ !<   »      , *»   *" * *4  f A %/ t


The Tubulous Boiler.


termed headers, which fit closely together. Every header is
connected, by a tube a, with a collecting chamber b at each end
of the receiver G; and all the tubes are expanded into their
respective sockets, the necessary holes at d being closed by
covers of ' mudhole' pattern. The water rises to the centre of
the receiver, which therefore serves both as dome and part of the
boiler. There is a cleaning hole at e. j is a cylindrical mud
collector, while K, L, M, and N are soot doors; and the draught
is compelled to follow the tubes, by reason of the division
walls c and D, the flame plates Q Q, and the position of flue E.
The receiver is held in place by the girders p P, bolted to
the brickwork. The headers are usually of cast iron, though
wrought-iron ones have been recently constructed, and plates Q Q,
with firebrick distance pieces, serve to stay and support the tubes
intermediately. The chambers b b are flanged and welded from
wrought-iron plate, the tubes are of wrought iron or steel, and the
receiver of steel plate. The flue may be at N instead of E.

These boilers have been much favoured recently by electric-
lighting engineers, on account of rapid steam-raising properties,
and immunity from accidents due to the small diameter of their
tubes, with relatively great strength; but they require considerable
cleaning and repairing. (See Appendices, pp. 828, 993, and 1061.)

Geometry required by the Boiler Maker.—This is not
of a difficult kind, but involves one or two intersections of solids,
and development of the contact line upon either ,of the solids
when their surfaces are laid flat He must know the relation of
circumference to diameter of circle, thus—

Circumference =

and TT •

diameter x ?r
3-1416 or ~

and the diameter of a boiler should be measured (for develop-
ment) to the centre of thickness of the plate.

The intersection of cylinder with cylinder is given at Fig. 315,
and the method of developing the plane surface: A and B repre-
senting a dome and boiler respectively. Taking the dome in
plan, divide the-circle, into, say, twelve parts, an<3 nwnber ;as
shewn. Calculate half-dome circumference, and lay out as at
CD, dividing into six parts by vertical lines. Project lines up

34-O                                     Geometry.

from plan to meet boiler circumference, and carry these along
horizontally to cross the vertical lines at c D ; the serpentine
curve, being then traced through the numbers obtained, will
represent the developed intersecting line. This may be repeated
on the second half of the plate, and allowances made for flanging
and welding. The boiler hole is developed by stepping-off the
three distances, /$, h^ and H, with dividers, and measuring them
from the vertical centre line in plan to give a, b, and c respec-
tively, the remaining four segments being symmetrical. The
length of plate is found by calculation.

Intersections of oblique cylinder with plane, or cone with
cylinder, are rarely required; but cone with plane is sometimes
necessary, as in funnels for American locomotives, or conical
flues such as that shewn at L, Fig. 310. The latter has been
chosen as an example, and the form of plate developed at K,
Fig. 315, J being the finished flue. The drawing j having been
made, the outer lines are produced to meet at f, and the dotted
circles struck, with gf and//as radii. Upon these are measured
the circumferences at d and e respectively, and allowance made
for welding and flanging.

If the set-squares at hand be not long enough, the marker-off
should be able to set out a right angle by the measurements of
three sides of a triangle, it being easily remembered that the
proportions 3, 4, 5, for base, perpendicular, and hypotenuse in
turn, will serve his purpose, as can be proved by the 47th proposi-
tion of Euclid's first book, thus :

32 +  42=  52      or      9 -f 16 =25

The length of arc, chord being known, is sometimes required,
and may be obtained as follows:—

Let *:=the half chord.
r=radius of arc.
a = half the angle subtended by the arc.

Then - = sin a.


The angle a being found from a table of sines,

Length of Arc = 2 x —- x 2w.r= '0349^.


/./"         /jjf


Set ting-out a Marine Boiler.

Setting-out a Marine Boiler.—We are now in a position
to detail the method of setting out the plates and putting together
of any form of boiler previously described. Taking first the
Marine Boiler: the Draughtsman must make a list of the plates
required, taken from the drawings, for ordering purposes, giving
each a marginal allowance, which will vary from \ in. to \ in.
all round according to thickness of plate. This is necessary, for
the shears at the Rolling Mill leave a rough edge and distress the
plate. Referring to Fig. 311, the plates N, P, Q, and s would be
ordered as 'sketch plates7; coming in roughly sheared to the shape :
G and x might also be cut down at the mill; but the remaining
plates would be ordered to the nearest rectangle. Care must be
exercised to remember flange or lap allowances. The Fox tube
is rolled by special machinery, so must also be 'ordered out'
When received, it must be carefully gauged at every ring, and if
found to be more than a £ in. oval, must be rejected.

Supposing all the plates have been received, we will refer to
the sketches in Figs. 316 and 317, taking the Front plates first.

I. Front Stay Plate.—This is received roughly sheared, as at
i. It is painted with whitening and marked off to drawing, as
shewn by dotted lines, keeping a near the edge to avoid much
planing. Then the curve b is cut out by band saw to give ant
edge for the flanging gauge 2. Flange to gauge, by the pro-
gressive method, Fig. 290, the ends being set as at 3, by the
horizontal ram. Being now considerably strained, the plate is
placed in a furnace, and uniformly, heated to a dull red heat; on
removal it is laid on a flat table, and straightened by wooden*
hammers, then allowed to cool slowly. The edge la is next
planed on the machine in Fig. 285, and a bevel given by setting
as at 4, the angle being i in 8; often this is'given to outer edges-
only. The long edge is planed with* a stroke the full length,,
and the flange 5 with short strokes, the position of the stops Qr
Fig. 285, being altered for the purpose. The Hanged edge is
milled as at 6, with a conical cutter, to obtain caulking inclination,,
a suitable table being provided to give a curvilinear feed.

The rivet holes are now drilled to the extent of one in "every
six, measured along the fitch line, for use in holding the plates
together while drilling in position. In this case the holes along


Front Plates.

s   II

! I

ia are to be marked from the tube plate; but those along the
flange are obtained by laying upon the latter a very thin steel
strip 7 prepared with marked holes, .and of the exact length of
the flange. After marking through, the flange holes may be
drilled in a Horizontal Drill, and the plate holes in a Radial
Drill. The holes for the stay bolts are marked off to dimension,
as shewn at u, Fig. 311, and drilled with clearance for the bolt.

II.  Front Tube Plate.—The Plate is first marked as at 8, and
part &a cut out by Band Saw.    The pieces ga are next drawn out
to a tapering wedge as at sl5 Fig. 311, after which the parts gfr
may be removed by Band Saw.     Flange gb to gauge;  anneal
and straighten.    Plane edges yd and ge to a bevel, trimming the
corners by chipping, and mill the flanged edge as before.    Set
out all the plate rivets as at w, v, and Y, Fig. 311, and the tube
holes.    Prepare a steel strip the exact length of the flange, and
pitch the rivets upon it; then mark through to the flange one
rivet in every six, leaving the corner rivets.    (N.B.—It should be
remembered that the corner rivets, where three plates overlap, are
always better drilled absolutely ' in position.')    Now drill all the
plate rivets under a Radial Drill, and the tube holes at the same
time, making first a small guide hole for the pin drill 10.    The
stay tube holes are made to tapping size, and the other tube holes
to gauge.    The flange holes are drilled in a Horizontal Drill, and
the stiffening plates (v, Fig. 311) marked from the tube plate and
drilled separately.

III.  Bottom Front Plate.—Whiten the plate as before; draw
centreline, line na, and strike curve lib.    Set out the centres
of the furnace holes, and strike a circle on each, smaller than the
flue by the flanging allowance.     Drill a small hole for drill
steadiment at the furnace centres, then lay the pkte on the drill
in Fig. 291, and take out the large hole by the trepanning tool 12.
Heat and flange, as at-13, each of the furnace Holes, and after
cooling lay on the marking table to test the» original lines, which
have drawn a little; so the curve i ib must be re-stnick, and cut
by band saw.    Flange to gauge, including the setting of the
flange ends; anneal and straighten.    Mark out line i id and cut
out with band saw; plane also the edge xxa.   If possible, give
the bevel at J when cutting, but if that is not convenient, finish

Back Plates.


by milling or chipping. Mark the flange rivets, one in six, with a
special steel strip, and the rivets along the seam a d, one in six,
from the tube plate. Set out the centres for stays Y YI} Fig. 311,
and niudholes z z, as shewn at 14. Next prepare the stiffening
plates 15 by marking out, sawing, cutting the oval hole by the
special method shewn in Fig. 291, and drilling the rivet holes.
Place the stiffening plates in position, and mark through all their
holes; then drill all holes by a Radial Machine, and cut the
mudholes by the appliance in Fig. 291. The edges of the furnace
flanges are tooled in the same machine by fixing the plate hori-
zontally on the table and revolving the tool Q Q, as at 16.

IV.   Top Back Plate is prepared in the same manner as I.

V.  Back Middle Plate.—This must be lined out as at  17,
with a and b parallel, and the curves struck.    The rest may be
understood from II.    After planing a and £, and setting out the
stay holes, the latter are left to be drilled till all are bolted

VI.  Bottom Back Plate (18) is treated in the same manner
as I., but the stay holes are all drilled in position, as in last

VII.  Front Ring Plates.—There are three of these, all equal
in length.    They are lined, as at 19, with long set squares, then
planed, the long edges to a bevel, and the short edges square;
next taken to the Rolls, Figs. 286-7, an(i Put through in the
manner previously described.    But many Marine firms prefer to
work with Vertical Rolls, believing that besides supporting the
weight, the curve is obtained more squarely with the long edge.
In finishing the short edge, a greater pressure is given to secure
accuracy of curvature, and partially avoid the necessity of bending
with hammer.   Now mark off the rivet holes to suit those already
drilled in the flanges of the Front and Back plates.    To this end
the steel strips are again used, and, being very thin, do not differ
appreciably in their outside and inside circumferences.    The
positions of joints T x must be found with relation to the butt
joints F F (Fig. 311), and the centres of XT marked upon the
front long edge of the ring plates.    Then the steel strips are
applied^ and the holes marked to correspond with the flanges.
Of course these strips must be all carefully numbered, to avoid


Ring Plates.

mistaking the one for the other. The rivet holes, one in six, for
the back long edge must be set out so as to bring the joints
F and B (Fig. 311) into exact relation with each other. B B are
therefore marked upon the Front Plate, and two methods occur
by which the intermediate holes may be traced: one involving
the use of the thin strips, and the other being the placing of one
plate upon the other, on blocks as at 20. The latter method
seems preferable, because all the holes may be marked on the
back edge of Front Plate, one in six drilled, and then traced
through to the Back Plate, VIII. The manhole is next marked
off, with-its rivet holes, but is not cut out till in position. The
butt strap is prepared by planing; heating and pressing to correct
curves between dies; then marking off all holes, but drilling only
three on each edge. It is next applied to the plate, these holes
marked through and drilled.

VIII. Back Ring Plates.—These are also in three, and of
equal length. They are marked as in the last example, and if
care be taken, the horizontal joints of the plates II. and V. will
be in line with each other. This is a necessity, so it is advisable
to keep the vertical centre line of the boiler well in view, on all
these plates I. to VIIL, during the whole of the marking off.

We may now bolt together the whole of the shell plates
through such rivet holes as have been drilled, and place the
boiler upon the cradle A A, Figs. 293-4, Plate XIII. The drill
spindle is adjusted as there described, and all the holes in the
ring plates drilled right through. There are two principal forms
of rivet holes required, as shewn at 37 and 38, the former being
for machine and the latter for hand-riveting. In 37 the arridge
is just taken off, while 38 requires 'a deep countersink, but both
may be given by the tools 21 (a and b). 2ia is applied from the
outside, and withdrawn when the hole Is finished. 2 lA is then
passed through from the inside of the boiler, and fastened in a
special slot as shewn. Its teeth cut left-handed, so the machine
need not be reversed, but the backward feed is given by hand,
and the depth gauged by a mark on the drilL All the shell rivets
are like 37, excepting those in the back flange, and even they
-y be machine-riveted, as will be shewn. The manhole is taken
: by drilling holes round its circumference close together, then

Furnace and Combustion Chamber.


finished by chipping. The bolts being clamped, their holes are
also countersunk, being first rimered to ensure exact correspond-
ence. The rivet holes both at front and back of boiler are next
drilled by placing the latter on a cradle, which allows the flat
plates to stand vertical, and face four drill standards supporting
horizontal drills on suitable saddles. The boiler joints being
truly level, the rivet holes may be easily drilled, as well as the
stay holes in the back, the latter being made to tapping size.

IX.  The Furnace Tubes (22 and   23) are  usually obtained
rolled, flanged, and cut to correct shape, an allowance being left
at front end for turning.     They may be flanged, however, under
the machine in Fig. 290, as shewn at 24, using special dies.   Mark
off all the flange holes, as at 23, and drill all those at £, one in
every six at <r, but none at the corners d.

X.   Combustion Chamber Throat Plate.—This is flanged to the
shape shewn at 25.    A rectangular plate being procured, the
centres of the furnaces are found as at 26, a hole trepanned, and
the flanging of the throat done at one heat, as at R, Fig. 290.
The rest of the plate is lined as at 27 and the corners cut, the
sides *, f, g, and h being flanged progressively until the whole fits
a cast-iron block or template.    This is of course an operation
involving great care.    Now the portions 2$a and 25^ are sawn
out, finishing the plate with the exception of the  taper ends,
which are drawn out by heating and hammering on the cast-iron
block.    After milling the flange edges, the rivet holes 23^, con-
necting with the Fox tube, are marked from the latter, and drilled
separately; and the flange holes carefully spaced out by reference
to the top corners and the furnace centres, but only one in every
six drilled now, and none through the taper portion.

XI Combustion Chamber Back Plate (28).—This must be
lined out and flanged progressively to fit a cast-iron block, and
the flange edge then milled. The stay holes are drilled in

XII. Tfae Cowr PLties for the Combustion Chamber are now
edge^planed, rolled, and beat bot with hammer, until thfey exactly
fit the flanged plates, as shewn in Fig. 311. There are three of
these plates, one for the roof, and one for each side; and the
holes already drilled in the flanged plates must be traced through



Riveting the Boiler.


' I' '

upon them. The inner laps at the joints b b b must of course be
tapered, but no holes are yet drilled there, or through any of the
tapered pieces.

Fix all plates of both chambers, including Fox tubes, with
temporary bolts; and, laying each upon its back, drill with
Horizontal Drill all the rivet holes, as spaced on the cover plates.
Mark out and drill to tapping size the stay holes in the mid cover
plate of one chamber only, and drill also the holes for the girder-
stay bolts. Set up both chambers in position as at 29 by bolting
through the rivet holes, and blocking below. Obtain level position
with great exactness, then draw horizontal tube centres by
squaring from the roof, and the vertical lines from the middle
plates. They are afterwards drilled to correspond with II. The
mid stay holes are marked from one chamber to the other by a
punch 30, of the same diameter as the tapping size of the holes,
and afterwards drilled by Horizontal Drill.

The Girder or Roof Stays are now cut out by band saw, being
clamped together, and are next fitted to the roof, as shewn in

Fig. 311-

The Band Saw is a very useful tool, but requires some
attention to keep it keen. The tool at 31 is a roughened steel
helix, rotated by gearing to sharpen the saw teeth as the band is

Riveting the Boiler.—The Front and Back Plates may
now be put together in a Fixed Riveter as *at^3 2, and the ring
plates attached by the same machine up to the condition L,
Plate XVI. But the Back Plate,must either be put in by hand
or semi-hand process, or by the machine at p, Plate XVI. The
combustion chamber (after riveting up) is first inserted, and laid
loosely within the shell. Then, if hand-riveting be used, the rivet
will appear like that at 38, the flat finish being obtained by very
quick consecutive blows from riveting hammers used by two work-
men, while a third ' holds up' a cupping tool within the boiler.
The hammering is continued on both sides after the rivet is cold,
as a sort of caulking. A pneumatic hammer is employed in
some works, as at 33, where a lever vibrates from a crank plate
driven by a belt, while the hammer end is provided with a
pneumatic dashpot or cushion, giving a finish like 36. The

J3,ojJjtr   jfizcj, 3/7


Screwed Stays.


holding up may be obtained as at 34 or 35, by pressing on the
levers there shewn.

But the boiler may be finally closed by machine, using the
methods at p or M, Plate XVI. The former is adapted to
internally flanged boilers, the tube plate being cut in three pieces
at the stiffening plates. After the flange has been riveted, the
various tube plate rivets may be closed by the usual Lever Riveter
with long arms, dropped in through the furnace holes. The best
result is obtained by a boiler designed as at M, Plate XVI.,
and this should be employed whenever the ship designers
permit it.

The 'Combustion Chambers are put together as at 39 and 40,
but the back plates are riveted by hand, with rivets like 38, unless
the flanges be made as at M, Plate XVI. The chambers and
furnaces are next put within the boiler shell, and the latter closed
They are slung as at 41, carefully blocked and bolted in position,
then clamped at the front. Placing the boiler on a cradle, before
horizontal drills, and on the machine in Plate XIII., drill the
stay holes through into the Combustion Chamber to ensure exact
alignment for the screw threads. All stay holes, including those
between the chambers, are now tapped, as at 42, by a tap whose
threads a and b are continuous.

The Screwed Stays are prepared on the machine at 43.
Stay a is coupled to spindle £, which revolves by gearing c;
screw b has the same pitch as die nut d, and prevents the forma-
tion of unequally pitched threads on the stay by 'drawing' or
uneven pressure. The stays, having a square on their ends, are
now placed in the boiler with a wrench, a nick being first turned
at each end to represent their exact lengths; so that having been
advanced to correct position, a sharp twist will break off the
surplus material. Nats are now added, and the stay ends
trimmed up.

The boiler being still upon its cradle, the rivet holes at the
furnace mouth are set out and drilled by the machine at 44.
The drill bracket may be revolved on a horizontal axis by worm
gearing, and this, coupled with the rotation of the boiler, will
enable us to drill all round. The riveting-up is shewn at K,
Plate XVI.



If il

|   At   1   4


4    i4


f*'   li^tit^
tn its*1

if*^* ^<«

f 4f»-               f||   J           ff^f

|  /,
ff >    »1               ,'         {f

«*}vrtt^oM4 |||r
* r   K fr» m$ flu

Pvff^ »% ^vvl M  llir           |4jl*\

*,y«r   |4 t»                    |«#               f|tc

*V                        III f **i*f»rf #

^ 4*^          0if

*I        «4jy|*ii|t^  t


^il              P         J        f^|i

»l          fll^ U|^       £


air                  tl <hr

ifirj^                     **ff §4                    nl

f|%r |*tf

|^|    |<A                             *i  f^*          p  4>-P                       r|fri

§ Jy?                                                »                                        *^%              l« * **l *rt$| «f 4

%                                                ofrrf              ll|

|#^                 1|,                  Iff                   r*|-|||f»|ft

j                    «4                                                    t»r

Hit-               *»**

1**   tl«C                                                                         41*"

*S|             tifi ill

|t*f                                                mf-p                                           Ih*?

< I ) if%                                  l^i            i* ii rr        ^ 1 1 i?f*

to            1 1           fM

tr * tt                              j^|i

JV, tt                i^l ^^0 "rwf«i*»  !^*

i>.# ^    #*f4;                  tit? <y f ^"ij*   >A

tf j* w§

''>   *«*  ffif*fr*f
'»t<*i- j*/p,ji tr

^ft  *^| 1^* | A? *

v f |X.M$            a  *

n*f|»^                     §f*f

f*fc^^I*fev$ i»^ ?*»*|^ *


4|f         */|           I*    |.ff

f-^tjif   fi^fc   I^Jl^   ^ |

«^|    %#*    f      t'f'1"   »*A

fc      I'/'t^/l.,       *^4       jr^/t   ^J^f

f !|    1*^1'   *     ', f    ^P*4?;^     I -

\«t    4>U     If*     t^.fy.,/    3*   «U*  «W,      *fi           5'*^     *  *   j{

,,     M   ,        7,        ..      „,       .,•,-...•,,,

^     "             ?   / i     * "        M ^   ^    t »               IJt'-'ll                   rl»»        i

j, «* x          j   5 ?      «  « «*                   '   *    ; *   '»        • ( i   f                                  |    t«

f $ *                              » / j t: '    ;                        •    •%    t n   ; (     I                     i «k(: - ^

4   ,                   * j                       ."                       y          "                           * /1      ,     (    j i             * i

#                        i             *                          '   <               /

,         J   ,     ?   *                  4^4

I              *  'f         '         , (          ',     ,   ,

Drilling and Riveting.


, All the plates are now prepared, and must next be marked off
for drilling. First the tube holes are carefully lined on the two
tube plates, and cut out by pin-drills in a radial machine. Then
the outer plates may have their seam rivets spaced out, and one
in every six drilled, always omitting the corner holes, or those
where three plates overlap. The various parts may now be
bolted together, and all the rivet holes drilled and countersunk.
Thus K and Q being connected, the tube-plate rivet holes may be
done in a radial drill; adding plate j, the circular seams may be
drilled, as described at page 308, including also the holes in the
dome-hole stiffening piece, and those for the smokebox plate.
The dome flange is marked from the boiler and drilled separately.
Bolting H to j, the firebox shell may be drilled round its circum-
ference in like manner, but those on the flat sides would be done
under a radial or multiple drill, the latter being preferable. The
barrel is now disconnected from the firebox shell, and the firebox
bolted to the latter; then the whole shell placed on the lower
table of the Multiple Drill in Plate XIV., and the stay holes
drilled right through both plates to secure accurate alignment.
All remaining holes are now made, such as those for the angles T,
w, and P ; for the seatings v and M ; for the palm stays at s^'and
for the guide stays at E.

The operation of riveting is clearly shewn at Plate XV. The
barrel and shell are closed by fixed riveter at A and o, and the
firebox partly by o and partly by portable riveter. Then the
smokebox plate and the firebox are each fastened to the boiler
shell by portable machines, as shewn at G, L, and H. Finally, the
dome may be riveted as at P, so there is no occasion for hand
work on any part of the boiler. Note that the angles w, T, and P
must be riveted before the firebox is put in.

The tubes are fixed by expanding at the smokebox, and beading
and ferruling at the firebox end, using the tools in Figs. 304 and
305 -, and the smokebox ends of the tubes are then cut off by the
tool in Fig. 306. The screwing of the stays will be understood from
the marine example, but in this case their ends are riveted over by
hand after fixing. The mudholes are tapped to suit the plugs, the
guide stays screwed into place, and the steam pipe M expanded
into the plate The boiler is lastly caulked throughout and tested.


Setting-out Lancashire Boiler.

The Lancashire Boiler (Fig. 310) may be next considered
shortly. The back and front plates are turned, trepanned, and
drilled throughout, with the exception of the rings a, £, #, and q,
these being marked afterwards from the angles. The shell plates
are prepared as before and drilled in position with axis vertical,
two by two. The angle ring a is also drilled for the shell, and
the holes at b for the flange; then all are riveted together in
batches of three, with a fixed machine, and the batches connected
by hand, or by the method at 34, Fig. 317. Next the flue plates
are rolled, welded, and flanged as at 24, Fig. 316; turned on
machine, Fig. 291; drilled in position by machine, Fig. 292 ; and
riveted together, with caulking strip between, by a portable riveter.
The plates/ and k^ are to have the angle rings/ and q attached,
but the plates themselves are first bolted to the other tubes, and
the whole tested with a long wooden lath to see if it will make up
to the same length as the boiler shell; then the end tubes turned
down accordingly. The general, straightness of the tube should
be tried during riveting, and adjusted by varying the thickness ot
the caulking strip. Now the rings of holes—a, b, J>, q—may be
marked on the end plates. First the holes at /, q, and a are
marked and drilled. Then the shell is laid horizontally, the flues
blocked up in place, the back and front plates put on, and bolts
put in the rings ay /, and q ; when the holes in the shell at b may
be traced through to the flange. Removing the, ba,ck plate to
drill the flange holes, the gusset stays are prepared with their
angles riveted on, and are placed within the boiler. The back
plate is dnce more bolted on, and the whole boiler lifted on to a
trolley, which can be run under a radial drill, the latter being
preferably hinged on a wall or shop pillar so as to be at a
sufficient height while presenting no obstruction beneath. The
holes g, A, and/are cut out by drilling, and those in the shell,
for the gusset stays, lined out by squaring from the end plates,
then drilled. Entering the boiler, the workman places the stays
in position, and marks off the remaining rivet-holes in the end
plates. Removing the back plate again, the gussets are taken
away to drill, then all are replaced for riveting.

The gussets, the flange b, and the rings / and q, must be
riveted by hand/' but the ring a may be done .by machine.


Girders and Ships.

Prepare the longitudinal stays and manhole seating; put in place,
with fittings; and test the boiler as before.

The Vertical and Water-tube Boilers present no further diffi-
culty. Taking the first, the shell is built-up separate from the
firebox and chamber. Machine riveting can be used for most of
this work " But when putting together, the foundation ring is the
only other part that can be done by machine; all the rest is hand
work. The tubes are expanded into the tube plates as before.

The Water-tube Boiler (f. 338) has its tubes cut to length,
and expanded into the headers; the chambers a b flanged and
welded; while the making of G will be understood from previous

As further examples of Plate Work, we illustrate a Girder at
Fig. 318 and a Roof Principal at Fig. 319; but these are'simple
in comparison with boilers, as far as their practical construction
is concerned. The Box Girder has its plates and angles sheared
to dimension, the holes then marked off, punched, and rimered
in position. The angles A and web plates B are first riveted,
and next connected to the booms c c: so it will be clear that no
hand-riveting whatever is necessary. The Roof Principal needs
no explanation. The first application of portable riveting to
bridge erection was made by Mr. Tweddell in 1873, on tne
Primrose Street Bridge, London.

Ships are now built of steel plates and angles, whose dimen-
sions are carefully got out by the draughtsman in the first place.
Much more drilling is now done than formerly, though a con-
siderable amount of punching prevails, and the plates are usually
sheared. The keel and framing are first erected, and the plates
then adjusted and marked from these. As regards the riveting
up, nothing could shew this better than the diagram at Fig. 301.
Of course there are many plates too long to be reached by the
machine, but this diagram shews what an extraordinary amount
of worf can be performed by these wonderful * Portable Riveters/




A     Area in square feet.

Bm    Bending moment.                       [efficient of discharge.

Modulus of transverse elasticity in Ibs. per sq. inch: Co-

' Larger diameter' in inches.

Modulus of direct elasticity in Ibs. per sq. inch.

Total stress in tons :    F° = Fahrenheit.

Total stress in Ibs.

Tractive force to overcome friction : in Ibs.

Weight of a cubic foot of water: Centre of gravity.

Height or head in feet: Total heat.
H.P. Horse power per min. = 33000 foot pds.
I       Moment of inertia -{2 (area x r2)}.

Joule's equivalent.

Modulus of volumetric elasticity in Ibs. per sq. inch.

Specific heat of a gas at constant pressure ; in foot Ibs.

Length in feet.    [~ Kv = ditto at constant volume.

Latent heat.        [_CP and Cv = same quantities in heat

Poissoh's ratio.                                                   [units.

Number of revolutions per min.

Coefficient of bending stress.           [in Ibs. per sq. foot

Total pressure in Ibs.: Effort, or force applied: Pressure













Ptons Total pressure in tons.










Concrete of formula for struts
Radius in feet.
Larger radius in inches.
Reaction at supports.



Water discharge
[in cub. ft. per sec.


Range of stress variation in Wohler formula

Number of teeth.

Twisting moment.

Greater tension in belt or rope.

Final temperature in heat mixtures.

Work put in.

Velocity in feet pec min.: Volume in cub. ft.

Weight or load in tons: Resistance, or force removed.

Number of bolts in flange coupling, cylinder cover, &c.

w ft
Concrete of formula for beam deflection •» —=p_

Modulus of section (in bending).
Ditto (in twisting).

Iff;                                                                J"


tft |t*| an** fi*
^H ni          f/j?

till        ill               ,*i* lint |**f i»t

*«* *|*»f lit

|f* AH #a
'%»jfc lit                       fff

t«fi tit                |»4

f                           4 1 i^» * *%* Hi

1^ i<>         •»

4|l    l^i           ^«^*| Aft

It* r,rV —



|| fit «*»li^i


Synopsis of Lettering.

/3 (beta),
y (gamma)

$ (delta),
il (eta).
8 (theta).
K (kappa).
M (mu).
r (//).
p (rho).
or (sigma).
r (tau).
<A (phi).
o> (omega).

Small Letters.

Coefficient of linear expansion in degrees Fahrenheit:

various angles.
Various angles.

„   .     r specific heat at constant pressure

Ratio of -£-—vj—-——-------- -   *     —-.

specific heat at constant volume

Deflection per inch length: £feet = ditto per foot


Angle of torsion.

Coefficient of jet contraction.

Coefficient of friction or tangent of friction angle.

3*1416 or ~: ratio of circumference to diameter.

Radius of curvature in bending: coefficient of resist-

Various angles.                                                [ance.

Absolute temperature in F°.

Angle of friction : entropy.

Angular velocity.


A      Total deflection in inches.
A*1    Total deflection in feet
2      ' Sum of.9


a    'Varies as/
>    Greater than.
<    Less than.

Parallel to; with fibre.
Across fibre.




OUR intention in this chapter is to treat of the cohesive
strength of the materials used in Mechanical Engineering, of
practical testing to obtain strength constants, and of the use of
the latter in proportioning machine parts, so far as may be done.

Load is the total effect on the structure of the external
forces, and may be ' dead' or ' live,' concentrated or distributed
(seepp. 391 and 438).

Stress is the cohesive force within the material called into
play to resist the load. (See Appendix IIL, p. 920.)

Strain is the deformation produced by the stress.

Kinds of Stresses.—Only three simple stress-strain actions
are possible: tension (pulling), compression (thrusting), and shear
(cross-cutting). Bending is a mixed action, and local compression
produces a bearing stress. Fig. 320 shews the distortions and
fractures produced by these various stresses.                        Ǥ

Elasticity is the property of regaining original shape and
dimensions after distortion ; very apparent in an elastic body,
but scarcely perceptible in a rigid one. In 1676, Hooke pro-
pounded the law * ut tensio sic vis' (as the tension, so this
strain), meaning that stress and strain are proportional, if within
the elastic limit of the material.

Limit of Elasticity.—A bar being subjected to an increas-
ing stress (of any kind), will receive also a proportionately increas-
ing strain (of the same kind) until the elastic limit is reached,
after which the strains increase more'rapidly than the stresses till
rupture occurs. Shewing this by a diagram, Fig. 321, o is an

Elastic Moduli;


origin from which stresses are measured along o A, and strains
along OB. E is the elastic limit and o E is a straight line, shewing
proportionality of the co-ordinates. Plasticity begins at E, and
(assuming a case of tensile stress) increases in perfection up to B,
the curve being interrupted at Y, the yielding or breaking-down
point (or commercial elastic limit), while the lowering at B s in-
dicates rapid contraction of sectional area at rupture (see A, Fig.
320). If the stress be compressive the material enlarges in
diameter after B is reached, and thus becoming stronger, the curve
rises thence instead of falling: neither is the yield point observed.

If W = load in tons at B, a = original area, and al = con-
tracted area :

W -5- a = stress per sq. in. estimated on original area.

and   W~tfx= stress per sq. in. estimated on contracted area.

The first is used commercially, and is shewn at B, while the latter,

the strictly scientific result, is given at BX ; and the plastic curve is

thus corrected.    The curve from B to s is not considered reliable.

Compressive stresses do not materially distort the specimen
up to B, so the curve requires no correction. The primitive
elastic limit occurs at E, after which a permanent set is given to
the bar. This limit may be altered artificially. (See p. 385.)

Modulus  of  Direct Elasticity, or Young's * modulus.
(E) is a number giving the ratio of stress and strain within the
elastic limit, and is practically the same for tension or compression
stress sq. in. in Ibs •'     /tlbs    /clbs
strain per inch length ~~  St  .     Sc '

Modulus  of Transverse   Elasticity,   or Modulus of
Rigidity (C), serves similarly for shear action thus:
shear stress sq. in. in Ibs.      j^lbs



C = -

shear strain per inch length
if will be understood by reference to Fig. 322, .being the strain
between two shear planes am inch apart.

Modulus of Volumetric Elasticity (K) compares stress
and diminution in volume, thus:

stress sq. in. in Ibs.            fvM

53 decrease in vol. per cub. inch      3V

* Dr. Tto* Young, Foreign Sec. Royal Society) 1826.


364                             Poissoris Ratio.

TABLE OF ELASTIC MODULI (units being inches and Ibs.).


Cast Steel Forged Steel ...
	30,000,000 30,000,000
	12,000,000 I3,OOO,OOO
	> 26,000,000

Steel Plates    ...

Mang. Bronze.,.

W.I. Bars W.I. Plates    ...
	29,000,000 26,000,000
	IO,5OO,OOO 14,000,000
	[• 20,000,000

Copper   ......

Gun Metal     ...

Cast Iron

Brass      ......

Muntz Metal ...

Water     ......
	,           -

Mechanical treatment may raise these ratios: for tempered steel
E = 36,000,000 and C = 14,000,000, while for rolled or drawn
copper E = 15 or 17 millions respectively. (See App> V^ p. 996.)
Poisson's Ratio (M) is a constant to determine the lateral
effect of direct stress. If a bar, as in Fig. 323, be extended or
compressed, it undergoes lateral contraction at A and expansion
at B. Then, within the elastic limit:

Direct strain == lateral strain x M
§t or Sc = 61 x M




Steel   ...        ...       ......

Wrought Iron ...

Cast Iron       ,..        „.        ...        .,, Copper
	a v

Brass  ...


Nature of Shear Stress.—If tfee bar in Fig. 324 be sub-
jected to elastic shear stresses, j s an equal pair of shear stresses

Nature of Shear Stress.



. 322.

by Pcwsoris JRatto   Jfyg* 323.


sl sl wiH be induced on the cube abc d (j>. 859). Taking the
corner <£, the forces Fs Fs can be resisted by the force Fc which
is whoEy compressive, arid the forces Fs Fs at corner e can be


Diagram of Work done.


similarly resisted by the purely tensile force Ft.    A force diagram
being drawn for each case,

Fc - fi Fs    and    Ft = ,/JTF,



Nature of Tensile and Compressive Stresses. — When
a plain tin-notched bar is broken by pulling, lines of cleavage
appear on the surface, inclined at 45° to the axis ; and the final
fracture is cup-shaped. Compression fractures are also inclined
at 45° and are often wedge-shaped. The evident deduction is
that rupture takes place on shear planes in both cases, and that
the three simple stresses are interdependent.

Work done by Uniform Forces. — The unit of work is
a foot-pound, or one pound exerted through a distance of one foot.
One pound acting through two feet, or two pounds through one
foot, are each two foot-pounds. Hence :

Work = pressure x distance

= pounds  x feet = foot-pounds.

These forming a product may be represented by an area, for
length x breadth == area, and A, Fig. 325, is therefore the diagram
of work with uniform force :

Work done = pounds x feet ~ o x x o Y ~ area A.

Work done by Variable Forces is shewn by diagram atB,
Fig. 325, As the body moves from ox to 5, the pressure varies
as GJ xp 2 &, Sic. Now, work done between o'j and i can neither
be ox KX x i ft nor i a x i ft, but must be the average of these,
or O-L /x i. In like manner the other dotted rectangles shew the
work between the remaining intervals, and their addition,
Area ox xx b vx « work done.

Work done in Deformii^g a Bar is found at r, Fig. 326.
Divide o B into ten parts, and erect pei^endiculars between the
divisions.   Measure the vertical ordinates in tons, then
Total of orflnates ^ ^ ^ .

10                                        '

me»n load x extension*^ work in inch tons. ($ee#. 1065.)




Resilience is the work done in deforming a -bar up to the
elastic limit. 2, Fig. 326, is the diagram, where B A is the maxi-
mum elastic load, and o B the corresponding strain. (Seep. 833.)


Work done =area A o B == c D x o B,

or generally,

Axy work within 1 = final max, tot, stress
elastlclimit     J      •-•.•    ••»-•.

straili =    ^


368                           Impulsive Stress.

Stress  caused by  Impulsive  Load. — When a body
moves with a given velocity, its store of energy (or work capacity)

= —- ft. Ibs. (see p. 98).     If this be absorbed by an elastic

material, we have:

work stored =    work given out

wv2         FIbs

-----     -. — x Aft (within elastic limit)

= total mean stresslbs x Aft (for all cases)


and              Total mean stress -. wv*

in Ibs.              2ffAft

which is applicable to steam-hammers, pile-drivers, fly-presses, gun-
targets, &c.

If the fall of a weight deflect a beam, or stretch a crane chain

work stored in weight ) __ J work done on material

«   .  _i n              > —. \    •           ,    . _,

in inch Ibs.


in inch Ibs.


x A

and Flbs is the greatest total stress, or the steady load which would
produce the same strain A.

Stress caused by Heating and Cooling.—Experiment
shews that the expansion or contraction by heat or cold of a bar

•  f       4        i



•'• '*/

of given material, is a regular quantity for each degree of tempera-
ture. When measured per inch length or breadth, and per degree
Fahrenheit, it is given in the following table : —

Heat Stresses.



Strong steel

„        „    tempered          ..... . Mild steel

Wrought iron ...
	0 /

Cast iron         .         . .

Brass ..............


	y j iii

Invar ...


If /<> == rise or fall of temperature, a /° = "expansion or con-
traction for every inch, and
Each inch is increased or decreased by a t° ins.

But strain by mechanical means is b <
Then if   at0

(Seep. 363.)




yibs ^ E a /°

and total force of expansion on walls, as in Fig. 327 at A B, is

FIbs= Eat a

Necessity of Testing to obtain Unit-strength Con-
stants. — It has been hoped that the cohesive strength of the
various materials might be obtained solely by chemical analysis,
but continued experience seems to shew more and more the
necessity for direct mechanical tests to obtain the strength per
square inch in tension, compression, and shear; hence the use
of testing machines. Certainly it is wise also to refer to chemical
oomposition in stating the quality of a material, in order to know
how far it is safe to heat or otherwise treat the same.

Testing Machines. — One machine generally serves for
tension, compression, and bending experiments, the pulling
shackles being changed to suit No doubt machines will ultimately
be designed to test all combined stresses, and thus verify the
theoretical formulae on which we at present rely. In small machines


Testing Machines.

the pull is exerted by turning a screw directly or by gear, but in
large machines hydraulic or other power is employed, while the
load is always measured by a smaller weight attached to a lever
or system .of levers, in steelyard fashion. (See pp. 834 and 1065.)

Cement Testing Machine.—Michele's machine will illus-
trate the above details, the load being applied by worm gear at B
to the specimen H} a cement briquette, and the pull measured by
the weight and lever c, or Danish steelyard. The arm D varies

very little, but the arm E increases to the maximum F, or some
shorter distance, during the experiment; the stress therefore varies
as this arm and the pointer is left at its furthest position after
rupture, while the weight returns about half an inch. The scale
is graduated to represent the full load upon H.

Horizontal and vertical testing machines are so named from
the direction of the pull, and each has its particular advantage;
the former Is represented by

The Werder Machine, extensively adopted in Germany,
and shewn in Fig. 329. c is the specimen to be tested, and B an
adjustable washer between shackle and crosshead A, to allow for
length of c. Ram D moves to the right by water pressure from
hand pumps, and the pull is given through the bolts E^ for
tension at c, or compression at G. Trie load is measured by the


Werder and Wicksteed Types.


weights j and lever H, the shorter arm of which is F, the pressure
being received on knife edges ^' apart (or much smaller than
shewn), and a leverage of 5bo to i thus obtained. A spirit level
is used to ascertain the horizontally of the lever H.

Professor Kennedy's Machine.—Messrs. Buckton & Co.
have made a machine to Professor Kennedy's requirements, em-
bodying the Werder principle with improvements. In Fig. 330,
A is the hydraulic ram in a fixed cylinder, and B a sliding frame
carrying an adjustable crosshead E, T shews a tension experiment
and c a compression experiment, the load being resisted in either
case by the crosshead F, and its effect transmitted throtigh the
rods GG to the system of levers. H corresponds to H in Fig. 329,
but a second lever M is here applied, with a jockey weight L to
avoid the trouble of changing weights. L is traversed by hand
gear at MX and carries a pointer at Q, while K' is a spring stop,
and j a hand gear for adjusting the position of *F by turning the
screws G G. In this machine all the operations are within control
of one experimenter and the specimen well in view; in addition
there is, during compression experiments, a shorter length of parts
between cylinder and weighing levers than in any other machine
(except the ' Emery'), as shewn by the thick lines in the figures
N, o, and P, thus giving less recoil on the knife edges at rupture.

The Wicksteed Machine, also by Messrs. Buckton, is
shewn at Fig. 331, as designed by Mr. Wicksteed to Professor
Unwin's instructions. A is the steelyard weighing lever, and B the
jockey weight, which at a leverage of 50 to i exerts 50 tons pull
upon the specimen. Additional weights up to ij tons at c exert
another 50 tons by means of 40 to i leverage. The knife edges
are shewn in detail at D, Fig. 332, being 20 inche^ long from
front to back; and the weight B is moved by screw #, either by
hand at E or by power at F, through the shaft b and gearing d,
the connection of the strap e being made immediately below the
fulcrum. Tfie levet is kept horizontally between stops H H by
admitting pressure Crater to the straining cylinder j through pipe
R, and the load is relieved towards the close of an experiment by
running back the jockey weight The pressure water is obtained
in Professor Kennedy's machine from the Hycbpaulic Power
Company, in a lat^r-built Wicksteed machine at the; Armstrong


JQO Ton TjteJtiria Mac/wrie

College by town's water acting through the Intensifier in Fig. 333,
and in the usual Wicketeed machine by means of'the * Quiet
Compressor' in Fig. 332* Crossed or open straps at /' drive a

Alternative Fulcra.


shaft k, connected by spur gear with nuts / /, which turn within
the bosses m m> and thus advance the screws n n. The latter, are
connected to the ram / by crosshead q, and thus a very even
pressure is given to the water, which finally passes to the straining
cylinder j, Fig. 331, through pipe R. The pump may be worked
by hand if necessary, or the strap fork moved by hand lever s if
power be used, and a cut-off gear at / puts both straps on loose
pulley when either end of the stroke is reached. The shackles
w and v, Fig. 331, are adjusted to suit the specimen by turning
the screws uu through the worm gear T; and x is to balance the

loose gear,1 from v downwards. Enlarged views of the shackles
are given at Y, Fig. 332, to clearly shew the gripping wedges,
slightly convex on the inside and roughed like a file.

Mr. Wicksteed's alternative fulcra, as designed for Professor
Hele-Shaw, are 'shewn in Fig. 334. Fulcrum A is employed for
heavy tests, and B for lighter tests, which are thus made with a
greater degree of sensitiveness. The lever knife-edges are level,
but the support c, which can be put in or out of position by worm-
gear, is higher than support D, as seen at (2). This gives enough
clearance for vibration either at (i) or (2), and the lever takes the
position E F when changing the centres. (See App. II., p. 836.)

The Emery Machine has obtained great favour as an
instrument of precision. Professor Unwin says of a 7 5-ton
machine: ' Every half-pound of load was precisely and: instantly
measured, whatever the stress the machine was exerting.' It is

rf; fit I**               *|f wf %                                  j#ff *lf

v pt ty^t *f                                      jfep ^                                        *tgi

jiujt           y4ff

JP%|  v>»                                                                 tftjjft

iwft^^iff                     M||          «          in ^               «%jf     <t If            «^f4 lf|,

^ffefc-.r;*                                                         V   ^   |* *|lf4^f!|l

w* *• ,/'^  f-H                          tf/*4'***'*'*• # *<*>tffsjfcft'i!1                  ^ II^A'I^ #            >

*     fc^   |**'||*vlif|*^    ^'jCf    '/M"S*|-f^^            ^'>^<rtjt   IV'ifift


Emery Machine^



M^uc/ujve,     Fjucr^ 335.

termed the c reducer/ and from thencfe to the lever weighing
apparatus. The movement of r is only -ooi", but the reducer and
support areas being as 1:30, the movement of piston s is

if I            *tf *

|i|           iff* 4 ft)** *                   tjyMtt  f                 *(f fin          i

* t

w                                    Ml    *f*  %J

jrt                    £'*                                  t        **f

»r*       I'^WfS                                                 J*f

«?-*fn        jty

»%f   <M'*fr«W''.*   *    4*<    ' l   W  ;   f ? * jfc&Ji 4^ rt f   |^|%'/V^^V                      "t    |**4/f    *14^>#%|

>'     4»*       ,    ,^/^ViA    r.f^^l^^^l    *    i/4    ||xA|i|,M*^s;^     *r,    |#»l»^|

».v >,«;** ; i *M^ll 'nK^ * W»g f'.tfy&'ta fH '*f|, |f?* * I I || "'I *»"| **"£*
'*/ I", >• « .j j*" ,»<»*''; 1'^$ if # HW<* A •»*»»*' * ^ 5|* *,*»r%Si ^||i t', |
#i|i t^ * » 'J , 4* / )**•** ***              *|^* t,^t«i^p f ifA^r * ^ ti "i %ti^*.,ff|»j|i

I                                                                                   %




^/*  I    '
/«f^*f Ik

1 I

Test Shackles.


to the specimen t>, held in place by a slightly conical ring c.
Compression shackles are shewn at Fig. 338. , AX and BX are to take
small specimens in a tension machine, and the arrangement at q

is for admission of large specimens in a compression machine.
The specimen at AX is placed at c, and the plunger d guided
within a cylinder. As one shackle a slides within the other


Strain Measuring.


shackle £, a very good axial thrust is obtained. Professor Unwinds
shackles at BX receive the test piece between a hard block <?, and
spherical surfaces d, and the parts are shewn separately to make
their construction clear. The Emery machine is provided, for
compression,- with spherical nuts A and B, upon which lie convex
plates or tables D and c, and the hard seatings E F receive the
thrust, c and D are adjusted to the specimen by means of the
handles jj. In the shearing shackles at Fig. 339 (designed by
Mr. Wicksteed for Professor Hele-Shaw), a knife A adjusts itself
so as to give equal pressure at B and c, while the specimen is
nicked down to localise the strain. The torsion grips at A, Fig. 336,
have sockets to receive a square bar turned down in the mid portion;
and Fig. 340 illustrates a pair of bending shackles where knife
edges B B are adjustable for various lengths of specimen, and the
shackle A is formed so as to indent the bar as little as possible.

Strain Measuring.—At first it was considered sufficient to
know the breaking load in tension, then Mr. Hodgkinson shewed
the necessity for compression tests, and Mr. Kirkaldy lastly
pointed out that the contraction of area at fracture was not to be
overlooked. Now it is considered imperative to know the
breaking load and elongation (usually given per cent., or extension
x 100), and advisable to obtain both load and extension within
the elastic limit. A stress-strain diagram, as in Fig. 321, will shev*
the whole life of the bar, and can be obtained in two ways: (i)
by noting load and extension at several points during the experi-
ment (the latter being measured by instruments of more or less
precision), then plotting a diagram to these dimensions; or (2)
by compelling the machine to make an autographic diagram.

Taking (i), the simplest method is to make a centre pop near
each end of the specimen, and measure the distance between
these by means of dividers; a better result is obtained by the
use of a standard rod c (Fig. 341), and wedge gauge D, placed
between clamps A and B on the specimen; and very great accuracy
by the aid of an exte&someter. Such an instrument is absolutely
necessary for the fine extensions within the elastic limit, and Fig.
342 shews a verf effective form devised by Professor Unwin. A
is the specimen to which Tee brackets c and D are clamped, both
of which carry spirit levels F and j, while D in addition supports
the measuring pillar G. Within G is a fine screw carrying a

fc**-^^                                    <^

^ ^




i %






f t * * * i * I*



11 i * i * i * -r

^. ^ 3 H        •   «   A  »

I « }



•*              ^t       **      #*            <*      ~f      ^T



i f 11 f =;; ^

Mi3**               -^     •     «fr

* % * I f -  **  ^ *

^ i *




* *


I * *; i: i - r

:fi|^rf :l

«   £- * 1 *   ?   r

-    I

Stress-strain Diagrams.


occurs.    This apparatus has been applied  in  Professor  Hele-
Shaw's machine.

Stress-strain Diagrams, as obtained principally by the
previous apparatus, will now be shewn (see Figs. 345 and 346).
The largest number of experiments have been made in tension
and our list of compression and shear diagrams is but meagre
In every case the authority has been cited, and where possible
the unit stress and length of specimens given.

Deductions.—Mild steel and good wrought iron have long
plastic extensions and considerable contraction at rupture (see
c, F, G, L), Stronger steels are less ductile, as at B and D,
while steel castings, A, are very short, though the strength may be
higher than shewn. Cast iron, Q, has really no elastic stage,
though Hodgkinson fixed an apparent limit, but brass, o, is
better off, and is much more plastic. N is a very fine diagram
for aluminium bronze, shewing great ductility and high elastic
limit. Torsional and transverse diagrams (s and R) are not
essentially different from tension in character, but compression
diagrams take quite a different form, v being a typical example,
the plastic portion tending always to curve in an opposite
direction to that of tension. T is an experiment on long pillars
held loosely in sockets to prevent bending; and diagrams Q, x, u,
and v have all been plotted.

Raising the Elastic Limit.—If the load be carried a little
beyond the primitive elastic limit and allowed to remain, say, for
24 hours, then removed, the bar will strain slightly -y but on re-
stressing, a new elastic limit will be found at a little higher load
than that just removed. Repeating the experiment beyond the
second limit, a third limit may be found, and so on until the bar
breaks. All this is beautifully given by diagram M, and also by
diagram s, one plastic curve bounding all the limits, and it is
clearly shewn why English engineers consider the breaking load
the only reliable test of a material. (See Appendices^ pp. 756
837, 1071, and 1074.)

Local Extension.—In Fig. 347 a test strip has been taken
12" long, and divisions marked across it at one inch apart, then
the actual extensions within each inch measured, and set up as
ordinates on the line A B ; c D E is the curVe shewing distribution
of extension, and is seen to increase very greatly towards the fifth



	i     ,

	•  /U
	\J   ft



	s: //v

	.5-    i,,.      i,.5-   ir
	3' 3-5


^    /


	/v -<Z/*

	/•5'        2








1U&H C/HI>(Jl.ArMI>



Double Test Diagrams.

inch, where fraction occurred. This indicates the necessity of
stating elongations somewhat as follows:—' 28*2 per cent, in a
length of 8'V or c 25*8 per cent, in a length of 10",' meaning -282
or -258 of the original length; and the breaking stress should be
measured as maximum load -~ original area. (See f. 837).

Diagrams shewing the elastic line have also been
drawn by Mr. Thos. Gray, of America, by means of the double
apparatus shown at Fig. 348. The paper drum is rotated by
worm gear, as in Fig. 343, to give the load, and there are two
pencils H and c, both connected to the specimen by wires; but
while A is connected to c through the single lever B and gives an
ordinary diagram, D gives motion to H through the triple set of
levers E, F and G, and thus the stroke of H is very greatly
magnified. Three diagrams are shewn, where the higher curves
are drawn by c, and the lower or elastic lines by H. Of course
two extension scales are required.

Admiralty Tests.—All war material must be tested as
follows, the data serving also as a general standard :—

	Tensile breaking stress in tons per sq. in. of original area.

	W. I. Ship Plates (1st class)   .,.
	( 22   ||
 ( 18 +

	W.I. Ship Plates (2nd  „  )   ...
	1 17  +

	W.I. Section Bars   .........

	W.I. BoilerPlates  .........
	\ 21   ||
 \ 18 4-

	Steel Ship Plates      .........
	26 to 30
	20 y.ms;;

	'   ,,    Castings (intricate) ......
	28 minimum

	„         „       (Roller Paths and Pivot Plates)
	36 to 40 : yield point |    at 1 8 min.

	,,         ,,     (Girders, Cylinders, and * Ordinary')
	•28 minimum

9 O
	Steel Rivets      .......... .,
	27 maximum

	,,    Forgings (general)  ......
	28 to 35
	28 to 2A °l in. 2'

^H  i 0
	,,        „        (Piston Rods) ,..
	32 to 35
	*.\j \.\j *i+   IOIM*,
 28 to 24 7 i& 2;/

	„        ,,        (Rollers and Roller Paths)
	| 38 to 45
	22 to 16 7 in. 2^

	„    Plates       ............
	28 to 32
	20 */a in 8"

	Gun Metal (ordinary)      ......
	14 minimum

	,,        (for hydraulics)
	3} f°0 in 2f>

	^ Manganese Bronze   .........
	25 9/l in 4"



<*WJLS> J.S'J^



	7    v






H    f|f!          IT                           *  **   ~   ^

**$ fr t | ^

f**                     ***   '*   *   '   "*

/ 4 *» f
A             n4                           «frt                                    ^ i|

n? W        ^                                         ** -y






f   /     \   V"*,*

f      f


f.   -      '    ^     %/»:      I    I, ^  f

*        I

r          V#*'4ff   *** IIV f 4 41 1^4 If f^All

V*   |f|   ff   4»|    f    f,


41 # *«•        i<«t
:^a4 ,   /j^/4

/f **  l| ^ f«f                          H/j  f«lf

la^ii /s- § /, ftit                  tt'ift   5 ^ r<*t

^1          tv                      I ' I                      //>

fti% 4*J luff i

•f                   HW                                     tlfll f4lllf  tuft

tlMP                                                  *rf »hr                                     It

I*                              *rl|                 «rf                              4ll4   *lii

llli                             1 4n          l«* ^|S                ;

iKf                     Iff* alt                                   *tt*

I/M4 <# Hfiil *ttm i                  ^ ^ <t| ^fr|f

If *                           ttf i Ir          t*»

«n4 to                        «Hr                                    Ufr


A tiw                       «                 ,    |#i-*%   |

*%                                        i             1 * $**'i

i * i

Praetors of Safety.

And the following table, deduced from practice, is fairly explained:

* Material.
	Dead Load.
	Live Load.
	Moving Load.

Wrought Iron and Mild Steel Hard Steel    ........
	3 i
	5 to 8 q to 8
	9 to 13
 I O tO I ^

Bronzes        ............
	0  ^ " 6 'tO Q
	10 to 15

Cast Iron and Brass   ......
	6 to 10
	10 to 15

Average Stresses adopted in practice.—We must now
sum up the results obtained in testing, as given by the best
authorities, and form a table of breaking and safe stresses.. But
as there are high and low qualities for each material, and samples
of each quality vary so much, our tabulations can only be the
averages of many averages.

Breaking Stresses.—Thus cast iron may vary from 5 to 15
tons per square inch in tension, 22 to 58 in compression, and 4
to 5 in shear. Wrought iron breaks at from 15 to 30 tons in
tension, and 10 to 22 tons in shear. The strength of steel increases
with the carbon it contains, but as a rule its elongation is
simultaneously decreased. Steel plates should have but J per
cent. Cementation steel reaches very high strengths, varying
from 40 to 67 tons per square inch in tension, some samples of
tool steel yielding 88 tons ; and tempering increases its strength.
Steel castings bear from 15 to 34 tons with reasonable elongation.
Capper depends on mechanical treatment. Cast, it supports 10
tons; rolled into plates, 14 tons ; and drawn into wire, 20 tons.
Brass has 8 to 13 tons per square inch tension, and gun metal
10 to 23 tons.

There is some difficulty in collecting good results for com-
pression. If the specimen be ductile it flattens out, and then,
as Rennie said, 'the resistance becomes enormous.1 Brittle
materials are more easily dealt with. Besides, tension has been
looked upon as a sufficient test for all materials, and thus the
compression and shear columns are in many cases vacant. In such
cases we may take compression = tension, and shear » 7 of tension.



*»      *   *-"

*   i   i ^
£   t   1*




^ ^.^ u* •<« ^^

T        i X

5   t

^   ^, ^*


«T *   »:  fc

1 ^ ^* ^

Classification of Stress Action.

The quantities in this table are given in tons, because the
numbers are thus more easily remembered, and because it is the
Engineer's language. Fig. 350 shews them diagrammatically.

(See Appendicesi pp. 840 and 1075 for further materials.)

The Proportioning of Structures and Machine
Parts by Calculation.—The equality of action and reaction
is the starting point in constructive calculation. Whether the
load be applied directly or through a lever arm, the external
forces must balance the internal stresses, and we have for the two

(Direct action).    Total load = Total stresses.

(Lever action).    Moment of load = Moment of stresses,*

which are our general strength equations.

Classification of Stress Action. — Practical cases of
simple or compound stress may be arranged under ten

* A moment = force x lever arm.

Tension Stress-Action.


Kind of Stress.

Some Cases.

I. Tension

2.  Compression ...

3.  Shear     ......

4.  Torsion (momenta! shear)

5.  Bearing............

6.  Bending (momenta! com-

pression and tension),..

7.  Bending + Tension

8.  Bending + Compression

9.  Torsion + Bending

i o.  Torsion -f Compression...

I Lifting rods, chains, bolts, ropes, boiler shells,
pipes and cylinders, boiler stays, flywheel

All short pillars.                                  Oims-

Punching and shearing, rivets,   pins, cotters,
coupling bolts, keys.

Short shafts, spiral springs.

Plate edges on rivets, cotter edges, and canti-

Beams, axles, boiler end plates, slide bars, teeth
of wheels.

Crane hooks.

Long pillars, boiler flues, ships' davits, con-
necting rods.

Long shafts, crank arms.
' Propeller shafts.

Tension Stress-Action.—Unit stress x area of section
will give total stress. Therefore :—

Load    =    Total stress.
W       -   ft a.

In the case of steam or water pressure the load is unit pressure x
area pressed upon, and

/tons x area of boiler end, or piston, in square ins. =/t a.

Of course/t may be either * breaking' or 'safe,' and W or p will
vary in like manner.

Example l.~Find safe load for following sections, at 5 tons per
•square inch, (i) 3 ins. dia, (2) 3 ins. dia. with f" cotterway. (3)
3" tube with 2" hole. (Eng. Exam., 1892.)

(1)    a--.

(2)     *-<

(3)   «-<




; tons.


=24*10 tons.

«/gg» 19*64tons.


*   ,

!                          ^4 * «f| #> I i

*> / •*                    py

^*   wftf^>*4t (»*|               f.f^ti^

^ 4 1

f   4    f     .      f   *   /      v,


^.'*'   f      ^   ^      *

^-      *      »          4

Strength of Chain and Ropes.                  397

The   Strength of Chain.—Both sides of the link resist
tension, so taking f = 4 tons safe :

but    r = -    .'. W = 6-28

2          --------------

4 x 4 = 25*12 rz
1 tons safe load.

Sir Jno. Anderson deduces a simple rule from the above :
(dia. in eighths)2

"Safe load in tons.

Thus an inch chain bears------= 6*4 tons.

,                                   i o               _

Strength of Ropes.—For white hemp ft — \ ton safe. But
as all ropes are measured by their circumference, and area = —~

Strength of hemp rope = £ x ~~ == "°4 circ'2 (tons)-

Wire rope has its members stated by their W.G. Referring to
page 276, the total area may be reckoned: then let/t= n| tons
safe for ir6n or steel.

Strength of Pipes and Cylinders, pressed internally.
Imagine a hemispherical vessel A, Fig. 351, hung by a string, and

pressed internally; then, as the vessel moves neither to right or
left, it follows that the total pressure on the curved surface in
direction F is equal to that upon the flat surface. The flat surface
is called the * projected area' of the curved surface.

Thin Cylinders.

First Case^ Thin Cylinders, — A boiler, or thin cylinder abed.
Fig. 351, tends to tear along the joints a b and cd. Examining a
strip i ''wide we obtain:

Internal load on
projected area

I — / sa^e strengtrl OI"two sections of
)      \          plate (in tension).

; x 2 rx i =ft x 2/x i.

/.   ptonsr = f ft

Suppose the plate tends to tear at a ring section as at <?/, then.
/tons x TT rz = /t x 2 TT r x t.

From this we find that


Stress on longitudinal section = —
and stress on transverse section = \ —

so there is no fear of a boiler bursting at a cross seam. The
above supposes the boiler plate to be of uniform construction
throughout. But as the seams, whether welded or riveted, are
much weaker than the * solid ' plate, a multiplier (?/) must be in-
troduced on the right side of the equation to reduce the quantity
and shew the strength at the joint. Then :

strength of joint
strength of solid plate
efficiency per cent = »7 x 100

) = efficiency =


For lap welded joints,
For single riveted joints,
For double riveted joints, 77 =


________________ jm^ tj      .* >roughly

For electric welded joints, 77 = -85 )     (See p. 755.)
and the formula for boilenor pipe strength becomes

Example 6.— A copper steam pipe 12" dia. is to resist a» internal
pressure of 160 Ibs. per sq. in. Find its thickness, if ^ for the brazed
joint = 80%

From above formula    - - x 6 == 2 x /x *8        .*.   / » '267 ins.
2240                     „           - •

Thick Cylinders.


Example 7.—Find the bursting resistance of a cast-iron pipe | in.
thick and 10 ins. diameter.    (Eng. Exam., 1887.)
/tonsr = // (there being no seam)

j0t°«s = •£- =-75 ton = 1680 Ibs. per sq. in.

Second Case, Thick Cylinders. — If cylinder thickness be small
in comparison to diameter, the stress at the inner surface is
practically the same as at the outside. But this is by no means the
case with very thick cylinders. Then the following formula must
be applied, devised by Lam<£ :

R"     D      V/t +/Tonl

"~                       •

and the stress varies throughout the thickness, the hoop tension
ft being found at any intermediate radius i by the following
formula:                                    ;*ton V2 <Y2 + R2^

A m tons =^2:;7^

Example 8.— A cast-iron hydraulic cylinder is 6" internal diameter,
and loaded with i ton per sq. in. pressure. Find (i) the thickness,
and (2) construct a curve shewing the hoop tension throughout the

•  R"=9     and_£=6"
r      V-/tOM           i'25-r                         y    - —

The section of the cylinder is shewn at Fig. 352, and the ordinates

at afrcdefg shew the hoop tension at the various rings, found as

follows:-                           1x9(814-9)

at a, yk= -~r )a7~~( « *"*$ tons
9    (oi-g)

Similarly  at £, /&**  76 ton,   at c«  '53 ton,         at ^= '406 ton,

at/«*283 ton,   and at^*— "25 ton,*


n '




The curve is an equiangular or logarithmic spiral. Large
guns are built of coils shrunk one over the other, so as to put
the inner tube in a state of compression. The pressure of the
explosion then tends to equalise the stress, by slightly adding to
the outer tension, but more than removes the inner compression.
When cold, a coil is slightly smaller than the core it is to envelop,
according to the following rule :

TX.   .     .        ,..,-.       meandia.2xr
Diminution of coil dia. =-

inside dia.


c for the outer coils =
c for next inner coils =
c for next inner coils =

= -00133
-• "00108

: '00083

Let an outer coil be 17" outside and 12" inside, then

x *°OI33
- ±*

T-V   •      •
Diminution =



The same effect is produced in cast-iron cylinders by casting with
a cold-water core, and thus much less thickness is required. (See
Figs. 289, 298, 299, 300.) (See Appendices I. and //., //. 757
and 841.)

< Casting Rule ' for Steam Cylinders, &c.— With the
usual steam and gas pressures, the previous formulae give so small
a thickness that the metal would not fill the sand mould during
casting, so an empirical rule must be adopted to enable the
cylinder or gas pipe to be cast, thus :

This will represent the thickness of steam chest and other parts,
but the cylinder body should be about J in. thicker, to allow
for reboring, and the flanges should also be stiffer.

Tensile Stress induced by Centrifugal Force.—
When a weight a/, attached to a string, is swung in a circular
path, it exerts a pull upon the string represented by the formula

>s.    (where v=actual velocity of weight)

In a grindstone or flywheel this centrifugal pull exerts a tension
between the particles of the material, which we shall examine in

Strength of Fly Wheel Rim.


Fig-  353-    R is the  average radius (radius of gyration) of tHe
rotating flywheel rim, a hb.    If«/ = the weight of a cub, inch of


the material, w h is the weight of the darkly shaded solid, and
its centrifugal force,

But every such solid in the circumference acts radially as at A,
Fig. 351, and the flywheel tends to burst at ab^ as the boiler did
at//, Fig. 351.

. •. Centrifugal force per "I          .   ^  ,             f  safe strength

sq. in of rim       \* projected area = { gtrip ^^ ^


^                ,.

and  /tlbs. =


2T**  2/lbsA

<   2   X   12 R         12

a b

then   ?7 = 1*64



For cast iron, w == '26 andylbs= 1*25 x 2240

«\   Safe ^ = 1*64 /Y •^ixJli?= X70 ft pej. sec


Strength of Bolts.

This velocity is reckoned for radius R, which for a flywheel may
be taken at the centre of the rim, b,ut for a grindstone

external radius            .

R __-----------------.._ .7 Of external radius.

A much less velocity (about 80 feet per second) is adopted in

* Strength of Bolts.—In an ordinary bolt with V thread,
the nut being deep enough, the bolt must break by a combination
of tension and torsion, '13 of the bolt area being devoted to resist
the latter, according to Unwin. In practice both are allowed for
by putting a small value on the safe stress—3 tons per sq. in.
for Wrought Iron, and 4 tons for Steel, estimated on the area at
thread bottom. Cylinder covers must be bolted very tightly, and
an initial screwing stress often resisted also, so the working stress
per square inch may be:

Steel bolts.
.. 4 tons .
.. 3 tons
.. 2 tons

W. I. bolts.
.. 3 tons
2\ tons
t\ tons

For 3 feet cylinders .
For 2 feet cylinders .
For i foot cylinders .

The diameter at thread bottom may be found from p. 213
and p. 192. Thus a f" bolt has a thread *i" pitch, and depth of
thread = • i x '64 = "064.

.*. Dia. at thread bottom = 75 - 2 x '064 = '622

22 X "'J 11^

and area at thread bottom =-------—   = '304

No faced joints, except very small ones, should have bolts
less than f" dia, or they may be broken merely by screwing up,

? 354.

and their pitch should not be greater than six times the bolt
diameter. In bolts that have to resist shock, the shank should
be turned down, as in Fig. 354, to the diameter at thread
bottom. (See Appendix If., pp. 833 and 842.)


'* I


.> 'j jf





I *H*

f ;



4 i«i

«i ^;|
» «I

* »

i I!




Strength of Suspension Link.


Example 10.—Find the thickness of a short, hollow, cast-iron
column of 18" outside diameter, to sustain a live load of 80 tons, plus
a dead load of 100 tons. (Eng\ Exam. 1888.)

Equivalent dead load = 100 + (2 x 80) = 260 tons
260 = 4 a       and a = 65 sq. ins.

But<2=7rR2-7r r2=65         or   — (8i-^2)=6s
andr= V 60*3 =7*8              .'.    /=9 -7*8 = 1 "z ins.

Shear Stress-Action rarely occurs alone, but pins and rivets
are thus calculated :            W = /s #.    *

Strength of a Suspension Link (see Fig. 355).—The
strength of one thin link in tension, at a and c\ the shear

strength of the pin d-} the strength at b] and the bearing stress
on projected area of e, should each equal half the load :
(i)           (2)                 (3)*      (4)         (5)

(w -

Let /t-i, /b=4,   and/,-}.
By (2) and (4)

By (3) and (4)       |x^~i*/

By actual tests      ^           =/

and* the thick link must be 2 / in thickness.



i Kbt      and w=:

and / =  '20 &
«. '66^

p. 415.

(See Appendix IV., #. 952.)



Strength of Riveted Joints.

Example n. — A wrought-iron suspension bridge chain supports
32 tons.    Find its cjiniensions and draw the joint to scale.


(£x'2£)      = 16 tons   and^=/\/j6~ _
~   '66x4   =   2'64             di

w — i '66 x 4   = 6*64
and the whole is drawn in Fig. 355.

t = *2 x 4 = *8

Strength of Riveted Joints. — A boiler plate may be
supposed to consist of similar links to the above, but with some
redundant material between (see Figs. 356 and 357). The joint






if H;1? L-.fa^









jEr&ruqth/ of-

may give way by (i) shearing the rivet, (2) tearing the plate
between rivets, (3) cross-breaking at ^, and (4) crushing by
reason-of too thin a plate.

Single Riveting; Size of Rivets.


In a Single-riveted Lap Joint, as n Fig. 356, shear
strength of one rivet = tensile strength of plate between two
rivets,                                v ^

or  f*---=/t(/


which is our general formula.    But the rivet (up to i" plates
bears a definite proportion to the plate thickness, thus :
d = 1*2 *J~i before riveting
d^ = i '3 Jt after riveting,   and t = *6 /x2

Also steel plates and rivets are the usual practice, where fs = 5 and
/t = 6. Putting these in the formula, we have

5 x 22 x d^         . „     ..         2
7x4             V         i        i

/.    pitch = i'09 + ^

which shews that the space between rivets is a constant quantity for
all plates up to i" thick. Also lap = 3 times d. (See Appendices I.
and Iff., pp. 760 and 920.)


Plate thickness.
	Rivet hole.


	•6 1







if '


i r>



	i "3


Efficiency: Double Riveting.

The Efficienc7 of joint has been already mentioned, and
its \alue,     .

__ strength of pierced plate _ p" — dl
strength of solid plate         /'

for single riveted joint,

with f " plate,   ij =


1-09 4-*8

= '57 or 57%

and    with i" plate,   rj = - f — =^45 or 45%

-            -

The Strength of a Double-riveted Lap Joint (zigzag)
can easily be discussed by reference to the * virtual links ' in Fig.
357. Clearly plate A must equal one rivet, while E equals two
rivets, in strength. So the centres at A will be 1*094- d^ while
those at B (called the pitch),

p"= 2 (1-09)+^= 2-


The distance from rivet centre to plate edge mil be ri </as
before, deduced from practice


: and the efficiency t\ =

|" platen = 2.*% + .s = 73 or 73%

Example 12. — Find the various dimensions of lap joints for f"
boiler plates; (i) ; single riveted, and (2) double riveted,


f and ^=-

and the joint is drawn at Fig. 358.

4* jtJ/ir *<p         t** 'i* * ? t   »f **•>«'» *t^t   f,( |   *               |               **|

*** 4 »^*it|              rf*p t«| *fijifff Jt»>i f*4$** j*»ji  t»                    4l»

f^            y|||                                                           |«

f 1 1 if®                f

*         *4i            4 *

i )wi»»                         -I ««   #i*            ^             |t

fc***l*j    |l^i|             ^^*i *«*M^    t"|   *'*'"''

tf^^|«l&          >n^>              t%«l/r-^^* f

t /*ri|«^f ** 'f t«*'p t/^l   t^|                         ^

mm»>«    ,*«^.'.«^

|f »  *« ; ;^ i —

fr           0


|j ...rr...-«



Chain Riveting\

may not be too large.    The efficiency will be reckoned on the
pitch line as before, because the joint is weakest along that line,

X"  ;  F HJ£ T

the links being most crowded there.    The fracture should always
be allowed to take place preferably on the pitch line.

All the. seams given may have two butt straps instead of a lap,
and the dangerous bending action be thereby removed, see Pig. 362.

Treble Riveting.




As each rivet is then in double shear, and twice the previous
strength, the pitch may be considerably increased, but this cannot
be taken advantage of, except perhaps in the thickest plates, or
staunchness would be affected.

A Double-riveted Butt Joint with two Coverplates
is shewn in Fig. 363, designed to use the full strength of rivets.
Of course the links, will be twice the width of a lap joint,
.•./' = 4 (

and   diagonal centres = 2 (1-09)+^= 2*1

The butt strap might be \t in thickness, but is safer at f /. It
might have to be thicker and should always be examined separately,
as one plate equal to the two straps put together. The overlap
may also have to be increased as at 2d.

The Treble-riveted Butt Joint at Fig. 364 is taken from


Examining as a butt joint, width of one link= '83:8 x 2 = 1*67 6"
and   /'c=5 (1-676) + 1*28 » 9*66"


marine practice. Using the general formula (page 407), (J>
= *%3&" j but as 5 links must pass at the pitch line :

Examining as a lap joint, width of one link = '838"
and   /' = 5

412                          Stringer Plate.

the intermediate value 8J-" having been taken.     Next, the butt
straps must pass 2-| rivets each at D1 and D2, or Dx + D2 must
pass the same strength as C.    But
Plate at C = 7'22/.
and* Plate at Dx -h D2 = 5-94 x 1-25 /= 7*4/,

or the links are most crowded at the pitch line.

Taking now the joint as designed,
Strength of pierced plate = (8-5 -1-28) 1*28x6=55 tons.

C X 22 X 1*28 X 1*28

Strength of rivets = J-----------------------x 5 x 2 = 64/34 tons.

---------------------                j x 4

Strength of solid plate =8-5 x 1-28 x 6 = 65'28 tons.
and    77 = -7.----~ ••

The Tie Bar or Stringer Plate, Fig. 365, is art important
deduction from the last example.   By compelling the joint to break

'. 365.

preferably at AB, the plate is only weakened to the extent of one rivet
The strips must not be bent abruptly, however, and the butt straps
should always be examined separately, and their thickness in-
creased until the links are narrowed sufficiently foi all to pass;
thus |f is required in the example.


	Dia. of rivet hole.
	Single-riveted lap.
		Single-riveted butt, with 2 straps.
		Double-riveted lap.
		Double-riveted butt, with 2 straps.
		Treble-riveted with i strap.
		Treble-riveted with 2 straps.
 = i rivet.
	w" == 2 rivets.1
	w" — 2 rivets.
	w" =4 rivets.
	«/' = 5 riv.       *
 = 10 rivets.

	5*45   j  '88

	5'45       "87


	i '09
	4*36     j   -82
		5'45   ;  '85

	i '09
	5'45   1   '84

	5'45   j   -83

	i -08
	5 '45       '83


	5 '45


	10 9




	i "3

	i '3



1     I



Remarks.—In cooling, the rivet exerts great grip on the
plate, giving frictional strength to the joint, but caulking
diminishes this, so it is not allowed for. Rivets over 6" long
.would break in cooling, so must be hammered up cold.

The formula for boiler strength, j)tons r—jt 17, can now be used
to better advantage, Construct.a table shewing t and?; tinder
all conditions, and after finding t x 77 from the formula, choose
such values of each as will meet the case when multiplied. Such
a table precedes this page, where the pitch has been taken at its
theoretical value; but must" be decreased to secure staunchness
where necessary, as with the thinner plates in the last column.
The efficiency table is not from practical tests ^ but jrom p, 4.07

Example 14.—A steel Lancashire boiler 8 ft. dia. is loaded with
. i oo Ibs. per square inch. Find ^, and indicate the joint yon would
' use.

p^r-At*   .: , „ = IOOX4^2 =-357
2240 x 6        JJ'

i. Single riveted lap joint         t
	rj = 2 x -49 = -367 4

2. Single riveted butt joint     )

3. Double riveted lap joint    \
	,= T?x-69=-382

4. Double riveted butt joint      /
	„= Li «3 =.363

Something between (3) and (4) would have to be adopted ; say
(4) with -J" plate and spacing like (3) for staunchness.

Example 1 5. — Two lengths of mild steel tie rod 7"xi"are to be
connected with double butt straps. Determine dimensions and
efficiency. (Hons, Mach, Constr. Exam., 1893.)

^=1*3.   Centre to edge =1*5 xi -25 = 1*87 5
vu" for one'rivet, in double shear = 2 x 1*09=2*18

7-1-3 = 57 width of pierced plate :
Checking we have :
Strength of rivets = 2 x

. • .         =3 rivets say,

^ X5 x 3^

Strength of pierced plate = (7 - 1 -3) x i x 6 =34*2 tons



Strength of Cotter Joint.

(5) Strength of solid rod = 7 x



(22                 \       II

— d1>-<l1t)=—<ll*-'

(7) Strength at b^ for shear = 5 x 2 x £2 ^    =10 b<idl


Equating, we obtain:
By (2) and (5): 14 <V = ~r^2


...... (8)

^- 7 ^ /==

By (2) and (5):

.'. /= '322 ^


By (i), (5), and (10) : io^/= — ^   .-. ^=171


-^ ,. Dt-..-44rfqr.4   (13)

By (7) and (5) : 10

The values at (12) and (14) are both unreasonably small, and
are increased in practice to DI«=2 d-} and <$l=^ = |^= 1-28 ^L

Strength of Shafts.


Example 16.—A foundation bolt with square head (Fig. 368) is
secured by a cotter. Find D, b, and / in terms of d> where ft,fs, andyi
vary as I : | : 2 respectively. (Hons. Mach. Constr. Exam. 1886.)

Following the previous calculations :

D = ro8 d,   b = 1-44 d^    and t—"*f>z d.


Torsional Stress-Action unallied with bending occurs
only in very short shafts. In any case the two actions Ihust be
separately considered. Fig. 369 shews a shaft under twist, the
external load being caused by the couple dxb c, while the in-
ternal resistance of the shaft is shewn by the couple exfg*

.-.    External mometitjlmoment of resistance of section
or    W>=/sZt

where Zt, the modulus of section, is a number depending on the
size and shape of the section.

Strength of Solid Round Shaft.—Let r be the outer
radius of a solid shaft, Fig. 370. Imagining the section divided
into concentric rings:

Total stress on outer ring =/s X2irrxt               (i)

Bntfs diminishes towards the centre because l>5 decreases:
.-.   Total stress at any other ring ^ = -±fs x 21?^ x t     (2)

* A couple is formed by a pair of equal and opposite forces, and can pro-
duce turning effect only, being represented by 'one force x "total arm.'

Round and Hollow Shafts.


The first formula may be represented by the lamina at a, and
the second by that at <£, and the total of the stresses on all the
rings will be given by the pyramid. Again :

Moment of stress at ring r = a x r
Moment of stress at ring ^ = b x ^

. *. Moment of all the stresses = contents of pyramid x average arm
== (base x \ height) x (f height)

.*. Moment of resistance of section =/s--------x - r =/s----

------------------------------------    /s     3        4        y    2

(and putting - = r)      = /s ( —7- J

Strength of Hollow Round Shaft. — Fig. 371 shews a
shaft of diameters D and d, externally and internally respectively.
At radius R" the stress is /s, but at radius r it is proportionately

less, being but fs


The strength of the hollow shaft will be
as stressed in situ,

TTD3    //• A7*"d*           TT /D4-aft\

"TJT "~ VSD/ ~i6~      ""•' 16 \     D    /

found by deducting the strength of shaft
from the strength of a solid shaft D.

Moment of resistance = mom1 of solid shaft minus mom* of core

(See Apj>. ///.,
/• 921.)

Strength of Square Shaft.—In this case we shall not
use the previous methods, but shall adopt a construction which,
although requiring careful drawing, can be employed for any
section, and is therefore general. In Fig. 372, A B CD is the sjiaft
section, divided into concentric rings as before. Erect perpen-
diculars on ED to represent the length of every ring, and bound
these by the figure E G L F D. j F = ?r s, and F E is a straight line,
while the lengths between F and D are found by stepping off each
set of four arcs with dividers. Now the stress will be greatest
at D, and will decrease gradually to zero at E, and the product
(f& x ring area) will proportionately decrease, so the total stress
may be obtained by imagining^ to be constant, and each ring
to have a value represented by the circumference decreased


Square Shafts.

according to its distance from D, the point of greatest stress.
Thus, if the ring K L be projected to D N, and N E joined, the
length KM will represent the virtual length of the ring if fs be
constant. Treating every perpendicular similarly, we obtain the
curve E p D, and the shaded figure is the virtual stress area, or
area of equal stress. Now cut out a copy of the shaded figure in
thin cardboard, and, hanging loosely from a pin in two -different
positions, as at w, mark plumb lines from the pin in each case
upon the paper. The crossing point will be o, the centre of
gravity, or centre of all the stresses, and the arm = E R. Next,
find the area of the figure. Divide s into 10 parts, and measure
everything in terms of these parts. Divide r> T into i o parts, and
draw horizontals from the middle of each part; then measure
their intercepts on the figure. Adding all these figures
(•13, -44, -82, &c.) and dividing by 10 we get the mean width
2*445, or "2445 s- The height DT measures 22*12 parts, or
2*212 s, so the area = height x mean width, and

Moment of resistance of section =/s x stress area x arm
=fs x height x mean width x arm

=/s X 2'2I2 S X -2445 S X -435 S      =/S('235*S)*

St. Venant shewed, however, in 1856, that the previous
methods (Coulomb's, ring theory) were not strictly applicable to
any but circular sections, and gave the following:

Moment of square section =^(-208 s3) or '88 of {^('235 s8) ]•

because the greatest stress occurs at the middle of each side. To
illustrate St. Venant, Thomson and Tait have imagined the shaft
to be a box full of liquid, which, if rotated, woukj leave the latter
behind somewhat, and the apices would cause two stresses—
tangential and centripetal—to act on the particles, the former
only being of momental, value.

* Generally Tm=^ -,   Zt =-,   and I=Zfc^, where I is the polar moment
of inertia (see p. 429),

Rectangular Shafts.


The Strength of a Rectangular Section was given by
St. Venant as follows :

Modulus of resistance of rectangular section = '294 4 -7^ — -
while the pure ring theory would give    *i666 b h ^W-^TP, and

the discrepancy increases with the ratio T
Thus if £=i"and h=*2n

Tm=*5266 ton ins. (i) by St. Venant; and 745 ton ins.

[(2) by ring theory,
and (i) — 7 of the diagram value (2).

If   £=i" and & = 4"
Tm = 1*142 (i) and 2747 (2) respectively
or (i) = '41 of diagram value.

Hexagonal shafts may be treated directly from diagram.

Strength  of Shafts by   Direct   Experiment. — The

following experimental figures may be used by way of correction.
Moment of any shaft d^ or s = Figure in table x (d* or s3).



Wrought Iron ..... .    ......
	6 -81

Cast Iron ........... .
	D Oj •\"?I
	uo 6*78

Steel ..................

Yellow Brass  ............
	2 'AC
	^*i <

Cast Copper   . . .....
	* *rj
	o A 0 2'74


A factor of 10 is to be used for short shafts and of 16 for long
shafts, to secure stiffness. Strength is rarely the sole criterion.

Example 17.—Find the relative weights of two shafts of equal
strength; the one solid, and the other hollow, with a hole half the out-
side diameter. (Eng. Exam. 1892.)

Strength of Coupling Bolts.
Moment of solid shaft = moment of hollow shaft

Let D = i.    Then d±= ^J
Weight of solid shaft

« -979

Weight of hollow shaft ~ 7r(R2

= 1-277 :

Example 18. — Find the relative strengths of shafts : —

2|" round,   3!" round,  3" square,* and 5" x 2" rectangular.

Moment of 2\" round  oc


•1963 x 15*62 = 3*066  say 31.

3!"     »      a     „    = '1963 x 42*87 = 8.415   say 84.

3" square  oc -208 s5 =   '208 x 27      = 5*616   say 56.

5"x2"rect.   a    -2944 -r-^-^      =5-467   say 55.

Strength of Coupling Bolts.—Fig. 373 is the face view of

a flange coupling.   As the bolts and shaft must be equally strong:
and the allowable stress on the bolts = |/s (see p. 415)

Moment of bolts = moment of shaft


/. 844.)

* >         >t

r*" ^-

tit     %

i I    ft       -   :      '      .• A      f."      //

'   I

Angle of Torsion.

Moment of shaft=moment of key (bearing)=moment of key (shearing)

. hid                         „ , ,d




Then by (5) h = '75 *'

By (3) and (i)    ^x-i963^=^
4                   4

By (i) and (2)

Let b = -3 t
By (4)   ^/=*2944^2        and l=r$d

In practice the following rules are adopted :
b = j d -f- ^r" and /^ =

Then /. '2944^

.-. >%/=-2944^2    (4)
••• *-|^.........(5)

Angle of Torsion, or the angle through which one end of
a shaft turns relatively to the other under a given stress.


p _ shear stress __/slbs
shear strain      6S

Referring to Fig. 376 and putting 6 in circular measure (viz. —? )

\      rad. /

^s                                                                   *         /lbs

8 = — for every inch of shaft length.    Substituting y-~- for bs


If a weight a/ produce a twist 0 (Fig. 377), then
utwr2=fs s_

Strength of Helical Springs.




oe referred to radius r (Fig. 378) if w be increased.
Strength    Tm a d*, while stiffness    ~ a d*.

&i.—-The angle of torsion of a round W. I. shaft is to
^ for every 3 feet of length, and the maximum stress is
•- per sq. in. Find the one diameter to satisfy both con-
rxs. Mach. Constr. Exam. 1889.) (See App. K, p. 997.)

22X2       2X8000X36           .          „

-——_.   or ——-, —   -----------^—  ^ d^—^'i&i

O d          7x360    10,500,000, xd       ——

'f ** <3L1 £&m—The angle of torsion being limited to one degree
IO feet of length, find the diameter of shaft to transmit
*t K£O revs. perm. (Hons. Applied Mech. Exam. 1892.)

22X2       2X/XI2O

7x360"   10,500,000 xd


^3 =


and /= 763*8 <f

763-8^ =



of  Helical (Spiral) Springs.—In the round
.(Fig-.   379)  the   pull   is  exerted axially,  and   any
f   if*    In   torsion.


Extreme elastic stress for steel =89,000 Ibs.

_        . .               89,000             .,

oxid. working stress = — ^ - = 29,600 IDS.


«|tiatre-sectioned steel (Fig. 380),








Deflection of Helical Springs.

Deflection of Helical Springs.—This may be found by
imagining the wire uncoiled, and treated as a straight shaft.
Let / = length of wire from A to B, and n — number of coils in
that length (Fig. 379).

cf. 380.

n,    andA=

2 /slbs lr


N.B. — This is for round wire only.    For square wire,



and for rectangular wire,           wnr*

_____         • //,/. 845.)

The above formulae have been thoroughly tested for steel with
C = 12,000,000 Ibs. and found reliable with that value. The curve
of work during extension or compression is found as for a bar
(page 367).

Bending Stress-Action.—Fig. 381 represents a model
devised by Prof. Perry to shew the stresses occurring in a beam.
Supposing W very heavy and beam / so light as to be negligible,

Ktnding Stress-Act ion.                       427

W causes a bending moment or turning effect round A equal
to W x /, and also exerts a downward pull to be balanced by
weight Wl;so that\V« \\\. The latter is called the shearing
force, and h felt on every vertical section of the beam. Wz and
W really form a couple with the arm /, and this can only be
balanced by another couple =(/ or t) x A n, a tension being felt
in the tipper fibres and a compression in the lower ones, shewn
respectively by the link A and strut B.

The case we have examined is that of a cantilever or overhung

Fig. 382  shews a supported beam  or girder, and the

action is here reversed, the lower fibres being in tension

the upper ft* compression.   Taking a bar of indiarubber, and

both before and after bending, it will be found that

€ is              shortened, t lengthened, while n is unaltered; * is

termed the neutral line or axis of the bar.
Position of Neutral Axis.—The bar "in Fig. $83 is beat
10 an            circle, and has A B far neutral axis, with fibres B c

€ G

Position of Neutral Axis.

in compression and BD in tension, The stresses will be zero
at B and increase towards c and D as shewn, forming a couple,
(j or F) >c G H, to resist bending, where j — F. Consider two
small areas at and a^ and let p = radius of curvature at neutral
line. Then:

Length before bending               = 2 T p

Length of ring at after bending   =2x1
Length of ring ac after bending   = 2 T (p -j
,*.  Strain on fibre at— 2 TT (p+-)'t) — 2 TT p   = 2 T yt
and Strain on fibre czc~ 2 T p - 2 rr(p-jc)   = 2 ?ryc


But A =J~ generally   . •. 2 Try =*J—


Total stress on a small area = /j = ±i^L"


E Tt ^t

Total stress on area BIDI = sum of —— for all portions of BL r>j


and Total stress on area BJ Cj_ = sum of   " °  c for all portions of \ Cj

But these are the forces F and j,

and as — is a constant,


SjK ^t=2j/c«c........................  (2)

or      Moment of tension area = moment of compression area.

But the centre of gravity of a lamina or centre of figure of area
is such that the. moments on either side are equal. Therefore tht
neutral axis of any bar passes through the centre ojjigure (or centra id)
of Us cross section.

Moment of Resistance.—^Again, in Fig. 383.


Moment of stress on a =

Moment of all stresses on a section

x y

P         J     P

But   2<3j/2 is the moment of inertia (2nd moment) of the
section = 1

"F T
.*.   Moment of resistance = -— (in terms of p)...... (3)

(Seep. 107.8.)


Moments of Inertia*




FOR BEAMS (rectangular.}

FOR SHAFTS (polar.'}





or Square














The same general formula
holds for shaft moment as
for beams,

thus Tm =/-

Note however the restrictions
at p. 420.



, ,,
_ -1- //» r






|f                     430                         Moment of Resistance.


1 p                         We may also represent the moment in terms of the limiting

!!                    stress/(sometimes/c, and sometimes/t).    Then:

if Jj                                       Bending moment = moment of resistance

j|                                                        Bm=/Z     ........................ (4)


I/                    and Z is known as the modulus of section,*


||                           Let y = distance of furthest fibre from axis:

I                     By (i)   /=           by (3) Bm = —      and by (4)    Bm=/Z

Z-—        .-.   Z = *...... (5)

p                   P        °                        y

and    Bm=/-                               .

_____I                              (Seej). 1079.)

The value of Z can now be found; thus,


% f

r                       i              •                   rr,      r "   '"        ^          f b frfi

rectangular Sections       /Z =/------r -   =/-—

and for circular sections    /Z = /—— -r- -

----------------------------£----      J     64         2,

Graphic   Solution   for   Moment   of   Resistance.—

Taking, first, a rectangular section (Fig. 384), draw the neutral
axis A B. Then c D will be the line of limiting (or greatest) stress,
and the value of any horizontal fibre E F to resist stress will be
found by projecting to c D and joining c D to N, thus obtaining
the intercept G M. Every fibre being thus treated, the sum of
the -virtual stress areas will be the areas c{# N and H j N, which
each make one force of the couple when multiplied by the
limiting stress f. K and L are the centres of gravity of the areas.

Moment of resistance (generally) = one force x total arm
Moment of rectangular section     = f\area C D N [• x arm K L

Unsymmetrical sections are treated at Figs. 389, 390 by this
method, which can be applied to any section. (SeeApp.IL>p. 847.)

•* Z = virtual area x arm.    (See Figs. 384, 385.)

Moment of Inertia, and Stress Area.


To find the Moment of Inertia of any Beam Section.

-Proceed as in the last construction and find Z.
nd I = Zjy.*    So for rectangular sections
bJP     h    ^bh*

62      12

Then Z = -



\   —
X   ---

Stress areas for circle, hollow circle, triangle, and hollow
•ectangle we shewn in Fig. 385, being measured as in Fig. 372
mean width x - Y The centre of gravity of each area is obtained
y cutting out in stiff paper and hanging up in two different

,|tf_   -    .....O —— —- —y

>ositions to mark two plumb lines, which will cross at G. For
he triangular beam the neutral axis must be drawn at ^ the
leight, a line of limiting stress drawn across the apex, and another
>elow, at an equal distance from its axis. Projecting and con-
urging, we shall obtain the areas shewn, which must be equal.
The results are as follows :
Moment of resistance of circle                =f'og82 d*

of hollow circle      =7-0982 :


of triangle


of hollow rectangle =7'1666:


* See also note, page 430; see also page 845.


#*   Hv   I ||   Jn**,                rflp  <i|^*                 !'*   |i^rip

>4  Jfe          ^»                                           '•«   4           . '.        %  /I/         **  •}

***£n            i, t  ffi 4                 i/^ff  rf*v   y  '«/'.*",«'*• >>

i       «*»    "4    *ft    '   Wl   frf ',14«|C      » «#'f I   *•*<">*!*  J»f      *'*l            »'fi

M i

' "(/    " 4# i   * /    f" a *.^    /• ; t

v      j*

I     /
f M :

Economical Sections.


Economical Sections.—It will be seen in Fig. 384 that
half the material is incapable of resistance on account of its
location near the axis, being only affected by shear, which how-
ever, has usually but a small effect. We are therefore driven
to the conclusion that ' solid' beams are uneconomical (seen also
in the solid circle and triangle in Fig. 385). The hollow circle
and hollow rectangle are an improvement, but the best results are
obtained by distributing the material near the line of limiting
stress, and thus the well-known H section (Fig. 388) is arrived
at for wrought'iron, where/c =/t approximately, while the modified
T section (Fig. 390) is add^ted for cast iron where fc >/t.

Assuming that the vertical web is for the purpose of resisting
shear, "we may find the moment of resistance by an

Approximate Method.— The direct strength of the flanges
forms a couple whose arm may be taken as the depth from
centre to centre of the flanges (the vertical web being neglected),
Let a = area of total depth of either flange,

Moment of resistance of H section =/cach   oifta^h

whichever is the lowest value.

In cast iron j = ^ or J roughly, and the flanges must have

/c     4
areas in inverse proportion.

Exact graphical solution may also be found, and we will take
a few cases.


Rolled Beam.


Momental Strength of Wrought Iron Rolled Beam

(the section being given at Fig. 388).—Referring every fibre to
c B or D E we obtain the shaded stress areas. As these change
in contour very abruptly, it is best to divide into 20 parts to
find the mean width '6895, Then "6895" x 275"= r8$6 area in
sq. ins. The arm may be found by calculation or by hanging up
the paper area from two positions, the first method being shewn
in the diagram, and the result found as 2*33" on either side,
Then Z = area x arm, and

Moment of resistance =/x 1*896 X4*66   =/8'&35
= 4 x 8-835 = 35-34ton inches

In such beams/t= 5 tons and_/c = 4 tons, so the lowest value
has been taken. By the approximate formula,

Moment —Jcach =4x1 '625 x 5 = 32'5 ton inches.

Momental Strength of Steel Rail (Fig. 389).—By
cutting out the section and hanging it, the neutral axis is found
at 1-56" from bottom and 1-69 from top; the limiting- line is
therefore BC. A second limiting line is drawn at DE, also 1*69
from axis, every fibre now referred to B c or D E, and the
stress areas obtained. Cutting these out, their centres can be
found, giving the arm 2*5", and their areas by dividing each
into 10 parts vertically. Then (mean-width x height) gives
, 751" * r69"= 1*27 for top area, and '867"x 1-56" = 1*35 for
bottom, area. It is very difficult to get them exactly equal
graphically, so the average 1*31 sq. ins. must be taken. Then
Moment of resistance =/x area x arm

= 6 x i '31 x 2 *5 = 19*65 ton inches.

Mornental Strength of Cast Iran Beam (Fig. 390).
—c D A B is the beam section, whose axis is found at E, Draw
perpendiculars F G and M NT. Set off H G = i J and M K = 4,
representing/i andj£ respectively, and draw Hjand KL through
axis, giving F j as 2§ and L N as 2^. This shews that if/c be the
limiting stress the tension flange would be stressed to z\ tons,
or dangerously; while/t at i£ tons would only'stress the com-
pression flange at 2§ tons, or safely, ft is therefore the limiting
stress, and A B the limiting line below E, while a corresponding

Plate Girder.


one P Q is drawn at equal distance above E.    Then as before,
after drawing stress areas:

Upper area = 1-256x4-56= s "7 1

T ri                   J         J      J ' v = 5*45 average

Lower area = 2-102 x 2-44.= 5-2 J                    °

Moment of resistance =/t x area x arm

= ij x 5-45 x 5-35 = 36-45 ton inches    •
or by approximate formula = i\ x 6 x 5^= 38-4 ton inches.

, --- ^_

Momental Strength of Plate Girder (Fig, 391).— The
rivets must  be  deducted in  tension  flange, but   they  aid the

resistance in compression,    The centre of figure (centroid) is then
found by taking moments of the parts round A,

Modulus of Rupture.


3 \


Moment of diminished section = moment of i rivet + moment of 2 rivets
(a - (£4- £•)}# = *5x arrrn-^xarm

_ (1-394* 8-5)-h(49'i6x 11-25)        „


-      68-28 -(13-94 -4- 49-1 6)      **

yc being limiting stress (see below), B c and D E are reference
lines, and the areas are found as before.

Each stress area = 26*35    an<^ arm = 2 x"12

For W. I. plate girders t/t= 5 and/c =4.

For Steel plate girders /t= 6 and/c = 5

The reduced jfc being an allowance for buckling.

.•.    Moment of resistance =/c x area x arm

= 4 x 26*35 x 3i '12 = 2226 ton, ins, for W. I.
= 5 x 26*35 x 21 §i2 = 2782 ton, ins, for Steel

Value of f in Beams.— If a < solid' beam be broken
.across, the ultimate stress, -deduced by applying the momenta!
formula, will be usualty- found much greater than ft breaking.
If then, the bending theory be pushed as far as the breaking
load, we must meet the case by the value

/o-0/t                        .

where /0 is the stress found by transverse experiment and called
the moduhis of rupture, while O we shall call the bending coefficient.
It varies with the beam section. Thus :

In sections ^ or  ^       O- is greatest, being about 2
In sections   |  or   H       O is less,       being about ij-
In sections *TT                  O = r

but depends also on the material, as seen in the following table
^compiled from experiment), and is often less than unit/ for woods.


Fir           |    -52 to -94

Oak         |      7 to i *o

Pitch Pine 1    '8 to 2-2

Cast Iron | 2; • 2-35;

Wrought Iroa | 1*6; • 175
Forged Steel | 1-47; • i'6

Gun Metal     |    i -o; • i '9
«•         ^ ) where ^=flange width

and &=wtb tkickness

And our beam formula becomes Bm = OfZ*

Bending Theories.                           437

In all cases of compound stress there appear to be points of
difference between practice and the theories adopted, Thus a com-
plete theory of bending ought to simultaneously consider tension,
compression, vertical shear, and hori2ontal shear, while the bending
theory treated on p. 428 deals only with the moment of the direct
stresses \ hence we should not be surprised to find the differ-
ences referred to. Some writers show that longitudinal shear is
exceedingly small as compared with the effect of direct stress

(as also is vertical shear, in beams where the ratio of length to
breadth is considerable), Others, again, assert they have found
theory and practice agree perfectly within the elastic limit of truly
elastic material. Without doubt the greatest portion of the dis-
crepancy is due to the attempt to use the beam formula during
the plastic stage, for which it was never constructed. Neither
-can it be correct to proportion beams by using a factor on the
modulus of rupture, the method largely adopted up to say 1875
or 1880, Probably a somewhat higher value of/ may be used
than/t (but not so high a. one as^), upon which to use the safety
factor im the case of solid beams ; while thin built-up sections
irtay be proportioned by putting the safe/t in the usual formula.
The greatest dilference between^ and fQ occurs, as we should
expect, with materials tliat have no |rue elastic stage, or, what is
the same thing, with such as have a constantly varying value for
E. Among these may be mentioned cast iron, india-rubber, and
some woods (see a ho p. 453).

We have now completed our investigations of moment of
resistance, and shall proceed to consider the left side of the
lending equation.

Bending Moment and Vertical Shear.—In long beams
ihe shear is small in comparison with bending stress, and is fully
met by the surplus section. For the distribution of shear stress


Bending Moment and Shear.

may be shewn to be parabolic, as at/s (Fig. 393), or greatest near
the axis, while on the contrary the  greatest  bending stress is
furthest from the axis, as shewn at A.
|/s x area of section = total shear

*~A f    s total shear load on section

ana js = -%---------------------:----------

area of section

In very short beams this stress should be considered, till
finally, in rivets and pins, the shear is almost pure. We will now
examine the distribution of Bending Moment and Shear Load
under various conditions of support arid load. (Seep. 922.)

I. Cantilever with Concentrated Load* (Fig. 394).—AB is
the beam and W the load, the latter having a leverage over

tJte= w

section A of W x / ton feet: at section D of W x f /, and so on.
The Bending Moments at various sections may therefore be repre-
sented on the base line ab by downward ordinates, thus :

AtA»W/xi      atD=W/xf

Atc=W/x|     atE=W/x^

and at B= nothing; and these ordinates are shewn at /^, hjr
and b.

The Shearing Force is caused by the reciprocal action of W and
Rt, and will equal W upon any section between A and B. For
regularity we shall always consider the force an the right side of

* Weight of beam is not taken, unless stated.



the section only, so here the shear ordinates are drawn downward
on the line a £, and equal W in every case.

II. Cantilever with Uniformly Distributed Load (Fig. 395),
the load being represented by the weight of the beam. Con-
sidering the beam hinged successively at A, B, c, D, and E, the
loads on the right may be successively concentrated at their
middle points, and the Bending Moments become:

At A-


4     ><S/    W/X32

W    /               r

At c=   —x-   = W/x -     cc

24        8

W    /              i

At D =   —x-   =W/x—    a

48         32

= 32


and at E = nothing, as shewn by diagram, and as the ordinates
vary as the square of the abscissae at a If, the curve is a parabola
with b as vertex.

Shearing Force on right of section A = W, at section B =f W,

W  W
and at c D and E, — •, — , and nothing respectively, as shewn by

diagram on a^br

III. Girder with concentrated load at centre and ends merely

supported (Fig. 396). — Reactions will each equal — , which we

shall use in. estimating the moment. Htl balances Rt2 round W
as a pivot and the stress at E is due to one or the other, but

not to both.    Then calling each reactipn — the Bmding Moments

will be :


-7    a

Atc= —




— x~



-     a 2

« 3



-     OCA


and similarly between j and E.

1 ..


r «*   \ -«


•»:         s««w»

»v         ^*

^          a           %        %

^         *         *        *•

•%          %          %         *

-   »-     e      w

t    i

-7     *        #


Girder with Uniform Load.                  441

A.        /W    7\       /W     /\                7

AtB=f~xr)   -f_.X-M     ^w/x-^-     oc    7

128            '

At c-f-^-xM  „ /:L._M     _ur/., 3


OC    12


\2     8V     V»" "i6V            128     "  ^

AtE = (7xO  ~GH)     = w/* I     - i«

and are  similarly found between H and E, all being shewn on
base ad.    Drawing/^-horizontally, the intercepts between kf
and tf/are seen to vary as in Fig. 395, and the. curve is therefore
a parabola with vertex at/

The Shearing Force is found by deducting upward and down-
ward forces on the right of each section.    Thus :


	= W-
	™1 _


 2   ~~
	3 8


	= ^w~

	2   ~"

	= ~W~




Ate-      ---

4      2.                4

-Wx|      OC-;

A.t J =       -IL==:^VVx-cx-4

2                      2

There is no force whatever at the centre,                     '..,.»/,.

VI. Beam fixed at both ends andloafad in the centre (Fig. 399}?.
weight of beam neglected. The beam will be deflected to the
dotted shape, A c and & j acting as cantilevers, and c G as a
supported girder. From c to G the Bm is upward, and from'1*


i *    '

Is       ^


9    **

s+^£ I

^ 4- - *

;*        3

v       i*   j-      ^

%       ^    *

5     !^ " * *

^                  ^     *

«- v »t *    ,

*- *

- r * i' * *4

:?? *^i


t    s


. Beams fixed at one End.


Case II), it will be found that the total ordinates under p q vary
as the square of their distance from q, proving that k q l^ is a
continuous parabola.

The Shearing Force at A consists of

•289 W+ -2ii W=—   and at c= -289 W

and the diagram is a straight line.    (See Appendix //., p. 848.)

VIII. Beam fixed at one end, supported at the other, and loaded
by its own weight (Fig. 401). — This case may be deduced from the
first figure on p. 850, and the point of contra-flexure is -25 / from
A and 75 / from J. Bending Moment

At   E = (-jr- J generally =

At ag^ -375 W x -25 / =
•25 W x -25 /

•0937 W/
"0312 W/

Shearing Force at A=:^ ^ 4-^ ^=8 -625 W, and at 1 = ^1^1
= -375 W, The curves will be found continuous in both
diagrams. (See Appendix //.,/. 852.)

Combination Diagrams are shewn in Fig. 402 based on cases
already discussed. The final shaded; areas Bfand Sf are found





444                  Combined Bending Moments.

by superposing the results due to the separate loads, having regard
to the signs + and -.    The cases are as follows:

(i.) One distributed and two concentrated loads on cantilever.
(2.) One concentrated and one distributed load on girder.
(3.) Maximum diagram for rolling load.
(4.) Three concentrated loads and one distributed load on girder.

The distributed load is more conveniently placed on one side
of the base line, and the concentrated loads on the other side,
in the superposed diagram. All are well lettered to shew the
relation between diagram and load. In Case (3) the load must
be placed over the successive numbers, and diagrams obtained for
every position, as in Fig. 397, then the bounding curve will be
the maximum, and the final maximum Em curve will be a parabola.

Fig. 403 shews a continuous beam on three supports, loaded
by equal concentrated loads, Wx and W1? and uniformly by its
own weight, W + W. The contra-flexure points are practically
the same for each case, and the diagrams can be obtained from
previous considerations (see p. 953). The following table is very
useful for continuous beams:—

Continuous Beams.



2 Spans 3 Spans 4 Spans 5 'Spans 6 Spans 7 Spans 8 Spans 9 Spans
	Appendices L and 7.L pp. 762 and 849.)
 3 8
 3 8
 32 28
	26 28
 32 28
 IS 38
 43 38
 37 38
 37 38
 43 38
	1 104

	161 142
	137 142
	143 142
	£37 142
	161 142

	374 388
 392 388
 386 388
 374 388
	1  I
 440 152 388 388

	209 530
 51* 530
 535 530
 529 530
 535 530
	1   1   1
 511 601 209
 530 530 530

Culmann's Funicular Polygon (Fig. 404) is a ready means
solving such a problem as (4) Fig. 402. Culmann of Zurich
proved that the bending moments are there proportional to the
:>:txlinates of a polygon obtained by hanging the same weights to a
uoose string hooked at the supports. Taking the loads in Fig. 404,
B o is drawn to scale, and represents the weights taken in order
shewn. Mark a point E any distance x ft. from c B, and join to
CL, M, N, and B. Draw from any point F, F o || c B, G H || M E, H j ||
EST 3E, and J K || B E. Join K F, and draw E L |j K F. The shaded
polygon is the curve of Bending Moment^ and

BM » vertical ordinate in Ibs. x x? (lb. ft.)


Cuhnanris Theorem.

Project ST from L, uv from c, w x from M, YZ from N, and
ab from B, and the curve of Shearing Force is obtained, measure-
able by load scale. Also s# —Rt and Tv = Rt2 (See pp. 855
and 1085.)

A Parabola may be drawn by the method in Fig. 405,
which consists in dividing A B and B o into an equal number of
parts, and joining the divisions of A B to D by lines cutting the
divisions of B c, then tracing the curve through the crossing

We will now take some examples to illustrate the equation of
Bmto/Z. The shear diagram is not often required in practice,
but should at least be made for trial in short beams.

Example 22.—The following beams are proposed for a central load
and given span : (i) a bar 4" deep by 2" wide ; (2) a bar 3*8" dia. ;
(3) a bar 3-5" square. What are their relative strengths ? (Eng.
Exam. 1885.)

Br^/Z   or — =/2        /. W oc Z



2 x i6


_22x 3*3 x 3*8 x'jS _


: 5'388    a roi

^ -    £L>L3J2i3^    __„.
"6                   6             ~

7*14.5    a 1-34

on Beams.


Example 23. — (i) A team 2' long- x i" square is broken by 250 Ibs.
at the centre. (2) Find the breaking load for a. beam 10 ft. long,
10" deep, and 6" wide, with the load 2 ft. from one end. (Eng.
Exam. 1886.)

For (i) (keeping- /in feet and bh in ins.)

^                _

"    4   ~"'    6   . •••~
For (2) by Case IV., Fig. 397





/                10

--—   and W-/*~ x 5 - 75Q x 6 x IQQ x 5
6           — y   6x8              6x8

= 46875 Ibs. = 20-9 tons.


r^Q. 406

Example 24. — A shaft pulley is 8' from one bearing and 2' from
the other. Weight of sliaft = 20o Ibs. Weight of pulley = 50 Ibs.
Total belt tension (downwards) = 100 Ibs. Draw bending- moment
diagram and find loads on bearing's. (Eng. Exam, 1888.)

t>        £   ^   r       '    L       W/       200X10            _,      f

Bm of shaft weight «= -g- =—5— =250 Ib. ft.
' .............. - '-•                            o    .         <>           ........ r   .............. -"

Reactions due to pulley and strap = — x 150= 30 Ibs.

and — x 1 50= 120 Ibs.
10      3

•"• Bm due to ptilley weight and strap pull =30x8 or 120 x 2 = 240 Ib. ft.
Jl^=ioo-f- 30= 1 50 lips.
Rt2 = 100+120= 220 Ibs.

And the diagrams are shewn .at Fig. 406, Bf being the combined
figure.            '     •,                                    •         i   - •

Examples on Beams.

Example 25. — Find the safe concentrated load in the following-
cases by the approximate formula.

(1)   Wrought Iron Plate Girder.— Each flange, 10" x J" ; angles?
3 i" x 3i" x i" ; total depth, 3 ft. ;  span, 28 ft. ; /t or/c = 5 tons.
Allow for |" rivets.

(2)   Wrought Iron Rolled Cantilever, H Section.— Each flange,
4i" x f"; depth, 8" ; overhang, 8 ft. ; ft or/c=* 5 tons.

(3)  Cast Iron Girder.— One flange, 3"xr|"; one flange, 9"xiJ";

depth, 12"; span, 20 feet.

"(All from Eng. Exam. 1891 and 1892.)

/ \   iTtr

(2)   W =                       - —   — -__

(3)/c <3C = 3 x ii x 4= 18 tons ; and/t at=9 x i J x i J= 17 tons

„,    4x17/2    4x17x10^-

.-.   W = 3 — '    =1 - i - ^tons      = 2-97 tons.

/              20X12                   - Z£- -

Example 26. — Find the depth of an engine guide bar 10" wide
and 4 ft. span. Total piston pressure = 25 tons ; length of connecting
rod = twice stroke ; and greatest obliquity supposed to occur with
guide block at centre of span. /0=5 tons.

(Hons. Mach. Constr. Ex. 1892.)

Draw crank and rod to scale, Fig. 407. Then the forces are as
at A, and the triangle of forces is drawn parallel, as at D.

_.    E     press, on bars        .,                .          1x25.

But — = *— : -   and press, on bars = - - = 6*25 tons
D      piston press.                                       4

4X10X5        -

Example 27. — The girder stays of a combustion chamber are 21
span, and are spaced 8J" centres apart (see Figs. 31 1 and 312), the sec-
tion being rectangular, 5!" deep by if" wide. There are two bolts to
each stay, 7" apart (Fig. 408). Find the greatest stress in the stay when
steam pressure = 225, Ibs, per sq. in. (Hons. Mach. Constr. Ex. 1891.)

The roof plate is equivalent to a continuous girder over three
spans, as at p. 445, so the pull on each bolt would apparently
be ^J- W. But the plate is continuous transversely also. Therefore

Axle Example.




Each bolt supports     ^J of \% W= i'2i \V
= i*2i x7x8^x225 lbs. = 7 tons.

Max. Bra at Bx or B2=Rt x 7" = - x 7 x 7==32§ ton ins.
and     Max. Bm at Bf=->°^-1-3— ~ 49


,   ,        49x6


Example 28.—An axle is loaded as in Fig. 409, with 5 tons at C.
Find greatest Bm; Bjr tending to fracture each journal; and deduce the
diameters at these places, taking/0 = 5 tons. (Hons. Mach. Constr.
Ex. 1879.)

Rtl = ^-x 5=2*14 tons   Rt2= - x 5 = 2-85 tons
Bm at C== 2*14x4 x 12 = 10272 ton inches

.'.   Bm at A = ~ x 10272 =6*42 ton inches

4<>                               -----------------------------------

Bm at B = -~ x 10272 = 8*56 ton inches


Now in circular beams ~» area goes for bending, and — goes for
shear (seeFig.385, p. 431).

,*/ ff   *   M f

*f| ?                                         i*ft|

itt»  t<» | »ii tr *|»ii-| »i|   *j/v/«|

Ji**{Kf i »%.ic          ^j             u^

' I*"?

»> |/if |

I   f M

t i



#~x r?-

*t* JP


Resilience of Beams.




Cantilever with concentrated load
Cantilever with distributed load
Girder with concentrated load ...

Girder with distributed load............

Fixed beam with concentrated load    ...

Fixed beam with distributed load       .........

Beam supported one end, fixed at other, central load
Beam supported one end, fixed at other, distributed load


A   a

and stiffness   cc


b ffi    ""*----""—*"*'*   —   /s

and the practicable allowable deflection is, for cantilevers

and for girders TV Per

sPan-    (See Appendix II., p. 855.)

The Resilience of a Beam is equal to half the proof or
elastic load multiplied by the corresponding deflection (see p, 367).
For a girder with central load,

48 El
„•. Resilience =


w      w'fi


and   cc   ^ h I

The Strength of Flat Surfaces in Boilers is best cal-
culated by the Board of Trade empirical rule.

Safe steam pressure/

r          f


where j— surface supported by one stay, in sq. ins.
./=» plate thickness.

100 when stays have nuts and large washers.
60 ditto, but exposed to flame.
36 stays riveted over and exposed to flame.


Beam Examples.

Beams of Uniform Strength.—If rectangular beams be
proportioned to their bending moment at every section, the depth
or width will vary as follows, easily proved by equation :—

With (
constant <
breadth f




Case I.           Depth    oc parabola.

„    II             „        a triangle.

nL> IV.    „        oc two parabolas.

;,   V.  '          ;;        cc

' Case I.           Breadth oc triangle.

„    II.               „       oc 2 convex parabolas.

„    III., IY.     „       cc 2 triangles.

3,    Y.               „       oc 2 concave parabolas.

Example 29.—A beam of oak, supported at the ends, 2' long, 2'
broad, 2" deep, supports 400 Ibs. safely, at the centre, and its de-
flection is -06". Find safe load at centre, and deflection of a beam
of oak 16' long, 9" broad, 14" deep, (i) with ends supported ; (2) with
ends fixed. (Eng-. Exam, 1882.)

& A2                        W I3

Taking / in feet and bh in ins.   W ex —   and A, ex -—

Sample "beam, W   oc



. 400 x 5



TVT        1                     Tir            QX   I4X  14

Ne\v beam,     W1 oc   ~-----——- = 110-2

(i) Supported; W : Wl : : 4 : 110*2   and Wx =
11,020 x 4096



= 25,600

.*•   A :  A! :: 200 : 1828

,    .       i828x*o6        ft,,
and A^••


(2) Fixed; Bn(i) : Bm(2) : ;         ;

4        o

,'.   W2=r 1,020 x 2 = 22,040 Ibs.


Example 30.—A beam, of unifarrn section is supported at the ends
and loaded centrally. Find tne ratio of depth to spaa that the'deflec-
tion may not exceed j^W of span when,/== 8000 Ibs. and E =s 28,000,000.
(Rons. MacK Constr. Ex. 1887.)

Combined Bending and Tension.


A =

48El   -d-T-^


A =


8000 x/2

__                          -     ...... _   _____

48 A /El     6^E          1000     6 x 28,000,000 #

/_ 6 x 28,000,000 _ 2 1
h     8000x1000       i

Combined Bending andjjTension Stress- Action. —

Let the bracket in Fig. 412 support a weight W.    There are two

actions upon the section : bending due to moment W r, and ten-
sion by direct load W.    Then—


( i ) Bending action :   W r =/0 Z an d f0 = — -




(2) Tensile Action :   W**fta and/t = —

W   Wr

.:  Maximum tensile stress (on inner edge of bK) = Ft = — + -^

O=i  for  H   and  T sections, and  O«iJ or 2 for solid
sections.    (Seep. 436.)

Strength of Crane Hooks. — In these, theory and practice
are considerably at variance.    The following table is regularly


Crane Hooks,

used at Elswick, and has been well tested, the diagram being
given at Fig. 413,


	•      P.

	,,13 3fTT
	Ij 1

	X4              ITT
		1 •'} IT

»i   •
	T    K ITir

1-   '.           !
	4-4         2if

1'             1
	Si        3
	1  »


r ;
	5 11 '    3TT


il,:      :

	r*, '

	!"'"' '
	rj         Z|

	V ; ,
	Q 2 7
	4         4

	! .
	•'TF      2
	1 ft

	ol      '       »1
 -ff        ;         ZW

If '
	1 1

a ,
ifij  •
 Taking O = 2, we have,-by formula.,/. 453, in tons and inches,
	i        *                              21 tons hook : ^ = 9*28, /=5,
	j'                                                                              -         21          21 X
	• '              rv?&       ft*    v
					— 2 26 + 8 47 — -"-^ 73 tons.

	$ tons 'hook: # = 3*95,   r=2
					'97j 2) = i'32
	1   »
 M'              '                                          .      / '1                                                         • '    J
			_   5   H.5X2'<
		^ — 1*27 + 5 '62 — 6 9   tons.              *
				f> 'r\t*         T-'^rtvx
	t||-                                                             c> yo    * 6^ ^ *
	<v                                  i ton hook : a= 2,        r= 2
					'i5> Z=  -5
	P/ '
	i                                              .'.   /
			I          IK 2TC
 —       »L                        ^             ««rft^..»«rf
		— 2 65 .

	, */                                 .                           .  *          5 ^ -

	1                ...... ............ ... ........
Examples in Tension -f Bending.


Apparently, stresses of 10, 7, and 3 tons are experienced with
the given loads, if the co-efficient O = 2 be used. But if the
bending formula be true without 0, stresses of 17^, i2-|, and 5
tons are involved. Now the elastic limit for wrought iron is from
12 to 15 tons, so the bending theory would appear to be insufficient,
and further that engineers are still wise in designing hooks and
such constructions by reference to the breaking load, on which
a factor of not less than 6 is adopted.

Example 31.—A longitudinal steel boiler stay, 20 ft. long and 2"
diameter, supports a flat area of 15 ins. sq., having on it a pressure of
120 Ibs. per sq. in. Find the greatest stress in the stay due to its own
weight and the steam pressure. (Hons. Mach. Constr. Exam. 1890.)

Weight of stay <w — 20 x 12 x 3*14 x '29 = 218*5 Ibs.
Steam pressure P =      15x15x120      = 27,000 Ibs.
^        w I       . TT d?        ,   _     14 wl      _       .,

=/>-—    and/o^TTwT = 8342 Ibs.

m ~~   8
ft a = 27,000
Total stress =/+/>



= 8598

' 8598-1-8342 = 16,940 Ibs. = 7*56 tons.

Example 32.—A piece of T iron consists of a web 4" deep and
f" thick, and a flange 2" wide and %' thick. Compare its strength
under longitudinal pull for the two cases (i) with line of action
through centre of web depth ; (2) with line of action passing through
centre of figure of the T. (Hons. Mach. Constr. Ex. 1888.)

See Fig. 414.

Find neutral axis by taking moments round A : k=i75". Draw
lines of limiting stress and find stress areas.

Z = area x arm = '6875 x 3*205 = 2*203
a = 3 sq. ins.    and r = 75"

Case (i)   /max =—H


1          Case (2)   /max =  — :


' Streng3hT(25 "

or as i : 2 roughly

Combined Bending and Compression Stress-Action

is calculated by the same formula as for tension and bending, by
substituting^ in the direct stress.

Fairbairn Cranes and Shi/s Davits.

Example 33.—Fig". 415 shews a c Fairbairn' crane. Draw the curve
of bending" moment for all sections, and design a suitable sectioa at
AB, taking/0— 5 tons. (Hons. Mach, Constr. Ex, 1887.)

Bm diagram is given in Fig. 415, using1 centre line of jib as base
line. At each section the

Moment — W x horizontal arm to axis of section.

Section at A B can only be obtained by trial and error, and has thus
been found in Fig. 415.    Checking by approximate method :

Area of two angles, one flange, and )       ,       -
portion of web between angles     j           "'

4 = 384    and total area = 45 sq. ins.

r = 15 x 12 = 180". •
10    10 x 180

•22-+47 = 4-92 tons.

Ships' davits are similarly calculated, but their sections are
like that* of a crane hook, and the same precautions apply.

Strength of Pillars and Struts.—Although these fail
by compression and bending, the action is not so simple. Struts
having a length of ten or twelve times their diameter are reckoned
for direct crushing only, but longer pillars bend before breaking.
Euler* devised a formula to give the greatest load consistent with
stability, that is, beyond which, the bar could not restraighten.

v "

Let Q = —jj—.     Then the stable loads w are given in  the

following table:—

* Pronounced «Oyler/

Long Columns: by Euler.



			One end fixed, the other free

				w   i    p '"

	Both ends free but load guided
	= rQ

	One end fixed the other   free,   but load guided

				W~2      ">
			Both ends feed, and load guided
	= 4Q

				^ - 4    p
A factor of safety of«5 must be employed, and I can be found
either from table (p. 429) or graphically (p. 431). The neutral
axis for I must lie across the greatest width of section,

Eulef s rules do not compare favourably 'with experiment, so
engineers prefer Gordon's formulae, which are a modified form of
those ma.de by HodgMnsort from his experiments. They give



*  *. £

t «*

?  v


f     i ¥ ^i 5 *

; J £ f 1 | <


I     **r          *?     »     2

'        *


i   ? * * s 5 J

t T 3

r '« 2

It t


*•*    »j


,? *J» '»/ Vn «fi


j,f w.



,t*f    J        '        -   ^^r,

*   j?   *? f*<*if

|       *<t *'*"  I

*#$*   /$?                         J^||r  t$£  HJ%%Jf|%   ,Jff|f              J/»


.     I' IS* * f^

§ |ft,             l;|   j                  -

I     | |    M*-^  *>Ti

'i Sf||r                  f'i

<rtj                       ,                                 ,

! ^ll'fll*^             V|»|^|   f

«   I

*// /** /


"1        *   4


**   I ^

f      »' I %


»*                                                                                 iff   .*"*4       *fr**<*$«4"   |*U'|

rf'ft                                     |^tf^,is   *V|    %4P|

|f    |'|^'»   I*                  #*&,» r*Lv£     frm     tf,fw

j4i                      tfW

" I                V*  f*||    x'

tt           |« n

*5fj,                     1*1     4i4^^

^-^ ln*^

*( ^|t   *"*    |t

< ,i   '   >,     II

-.'tV   '4V


Combined Torsion and Bending.

is the equivalent Tm, and the darker diagram, half of this, is the
equivalent B^.

Example 35.— In Fig-. 419 are some dimensions of a crank shaft.
Let P = i ton when at right angles to plane ACB, .Tm being balanced
by a couple M at D. Find greatest Bm-hTm and diameter of shaft
when/; = 6 tons. (Hons. Mach. Constr. Ex. 1893.)

The end view shews how I\ must be introduced to .make the
couple P Px complete. Then. Pj produces a B^ of ^| x 12 = 5-45 ton ins.
as in diagram, and P gives a Tm of 5 ton inches.

equivalent Tm = 5-454- A/547 = 12-85



(Seefi, 997.)

1*0 *ft i                           *4ffr                   i

in             «»*»                   fit* #* f **•* *         t IM«J

*f+.j *yj ^                       ft          r //         */*#

ty           if ft*                                        I

t*i /i|
|r         r         i^^^l" **£*'&>         i/     t


1 1 ^

*,#*.' I «'   %.   f   -/. ^ Jf'^ «.'/#   -f"   /

*»?tl     J rle|  jsflt^     *  , r    «,       'I    -^     .   tp ^   /*"»"U   4'^  /  *      t«   I   C'i}   W       ^f ^


f    £ ,t        %t i -    I        ' -  "I    ,- i;^

j *^«««Mr^    ,
> ,   t    ?   »      J

,*'••         ,    *'

^   %•*!''             4''   If

M       t     *       ' .    ,'        t                i




, , • - V •'. ,



>f|    «|
*rf    |f    f^ift


i*4|| -i^f

*s#* *                     *4t*-n« ^  ';

J«|*»;    Ut   'f   *    *

t   4**|*l *|j|$tf   f|* ^pj/r^

i^i '^«vt  f ,v < , t',

/•   >//

1  *4               Jt'f |s        tf

3f*«v*j* *>»$

fef I/i c

—    *

T- l"" L •• *   i

f •• . •••*


I/" 4 »n

||!    ^    |t    J»-    .^

*   V     "«   J If


Suspension Bridge.

polygon does not properly close, the arrows may not be in the
right direction, and a new examination must be made.

Next tak,e the point <2, and draw the triangle F A, A E, E F in
the manner shewn at j. EF should measure 15 cwts. Finally,
'draw the polygon EA, AC, CD, DE for the point ^, as shewn at
K, making CD = 13*9 cwts. and D E = 10 cwts. And the stresses
in the members may now be measured off, marking -h for com-
pression, and - for tension:

-f in AF = 14*5 cwts.
4- in AE = 104 cwts,
- in AC = 9*5 cwts.

Thick and thin lines also represent compression and tension

Suspension Bridge Chain.—A free uniform rope or chain
hangs in a catenary curve, which is, however, so nearly like a
parabola that the latter is always substituted for simplicity in
practice. Taking the chain in Fig. 424, supposed weightless, but
with loads at even distances as shewn, the forces at L and B are
necessary to keep equilibrium, and the chains will be in tension as
shewn by the arrows. Supposing reactions to be 3^ each, the
triangles ABC, A CD, AEF, &c., are drawn in succession. Then
the distribution of load may be found for CD, DE, EF, &c., and
the stresses in the chain also measured.

Warren Girder with Symmetrical Loads.—Mrst,
Distributed on lower boom (Fig. 425). The cells are equilateral
triangles, and the girder has been much used for American
bridges.' Loads being i, i, i, reactions are i^-fij, and the
force i\ at• j causes compressions in H B and H A, but tensions in
AJ and AB. The force diagrams are drawn for points i, 2, 3, 4,
5, &c., and the total diagram is given in the figures JKGDA.
Measuring the latter, we find the stresses to be as follows :—

In A H and H G = 173 4-
„ B A and G F =? 173 • —
„ CB and FE'= 0-58 4-
„ DC and ED = 0*58 -

In HB and HF =  173 +

„               HD  =  2-31  +

„ AJ and GK =  o'86 -

„ CM and LE =  2*02 —

*< ft                                                       |f|                                                   «||

am                           IK/f

»                                                     if                JKJ jn ||      * |t

jrtft fi|f        \||  */   -iff           *,f|  */

ffl    *|tt                                                  JK*|                       4*f                |*|    ' *n!*§

,*    r *      j        *

*    * ^       *  *    >

^       •*   >    *       *
<*       *       *       ^

kw                 L

' \**/  f/ "/ " A**

4f     *


M f

,'r  Jr   {frii*|


Jib Crane.

gives stresses due to bridge-weight, and 427 those due to loco-
motive, &c. Tabulate the stresses so as to find the maxima,

					Maximum live load.
		Stress due to dead load.
	Total maximum.

	2nd position
	3rd position
	4th position
	. 5th position
AH HG BA &c.
	4. -f

After which the bars are designed to meet the stresses, either as
ties or struts. A lattice girder is shewn in Fig. 426, being two
Warren girders superposed.

Example 37.—The post, tie rod, and jib of a crane (Fig. 428) are
15, 45, and 50 feet long respectively. Find all the stresses, (i) with
barrel on tie rod ; (2) with barrel on jib, with a load of 5 tons. (Eng.
Exam., 1887.)


Reactions.—Load produces both turning effect and downward pull,
resisted by equal horizontal forces at b and c, and by 5 tons upward
force at c. That at b is supplied partly by bending strength of post,
and partly by balance-weight. Letter the truss.

First, suppose the weight hung from a as a fixed point. Draw
B C in stress diagram = 5 tons, and complete triangle B C A. Passing
to b} E A must be a pull to product equilibrium, and B AE is the force
triangle. Finally, forces at c are shewn by polygon A C D E A, and
stresses are :

onAB=i5-      on CA = i6*66-h      and on A E = 277 -

Redundant Members.                         469

Also E B = C D = 1474.
and weight couple = righting couple

Tons.    Feet.                 Tons.    Feet.

5X44'22        =         1474X15

22FI            =          22I'I

Case (i). Stress in tie rod from dtoa becomes - 15 4-5 = 10 tons —
Case (2). Stress in jib from e to a becomes H-i6-|4-5 = 2i|tons +

other stresses being unaltered. The advantage of Case (i) is obvious.
If the barrel be between d and e the stress must be resolved on the
two members. As the load varies from nothing to a maximum, the
righting moment of balance weight should be half the maximum.

Redundant Members are such as receive no stress accord-
ing to force diagram, but contribute usually to resist buckling.
In Fig. 429, cross-members connect weak strut B with strong tie A,

but otherwise receive no stress.    In Fig. 430 A is redundant, but
receives stress due to instability of strut. B.   (SeeApp. IIL,p. 923.)

Example 38.—A crane is constructed as in Fig. 431. Draw the
stress diagram for internal and external forces. (Hons. Mach. Constr.
Ex. 1888.)

The crane is shewn at a, and the stress diagram found at
mencing with the weight E D.

I"  *

m *»f%<

t*N •.:<>.«•




*   V     '

k,>;".....^   %>




J1*/ It


*  ^        * i,    U     ' f





Wind Pressure.


60 Ibs. per.sq. ft. (according to the exposure) upon the area
(k x width of bay).

Total force Pt = 56 x k wf,    say, in Ibs.

Then Pt may be resolved into two forces, one parallel to the
rafter, and one (Pn) normally, and

Pn total = ~—-                       ff

This force is distributed at a, d, and c as -fV .^n> f Pn, A Pn.
In iron roofs the expansion is usually allowed for by fixing one
end (a) and leaving the other free (£), which allows us to say the
reaction, Rt2, is vertical. Now Rtl, Rt25 Pn are three balancing
forces, and must meet in one point X, found by producing Pn to
meet Rt2 produced. Then joining a X, the direction of Rtl is
found, and the amounts Rtl -and Rt2 further obtained from the
auxiliary diagram. If the wind blow from the right, Pn is exerted
on cb, and x will be above instead of below b (Case III.). Both
II. and III. must be examined, though we only, have space for
Case II. Take the lettering as in L, with the exception of the
additional external space L, and draw the stress diagrams as
shewn below.

Finally, tabulate the stresses for Cases L, II., and III.: then
add L to II. and III. separately to find the maximum stresses,
and design.th'e members to suit. (See Appendix, IV,, p. 953.)

Framed Structures of Three Dimensions are such
as include a solid instead of an area. They must be solved by a
step-by-step process, taking each plane in succession. We will
explain by means of an example.

Example 39.—A sheer legs (Fig. 433) is formed of two fore legs
145' long and 60' apart at the base, and a back leg- 170' long attached
to a nut having a travel of 40'. The maximum overhang is 40', and
load too tons. Find the stresses in the members, (i) and (2), at each
end of the nut stroke; (3) when the load'is directly over the base
plates, (Hons. Mach, Gonstr. Ex.'1888.)

Case L Nut at P.—Taking first the plane AD B, the diagram P
shews stresses . A _,

Turning to the end view, the stress of 174 in AB must be resolved in
each leg, as in diagram Q, giving stresses

in a b and a c, each =* 88 tons.


Three-dimensioned Truss.

Case II.   Nut at E.—R is the first diagram and S the second
giving stresses

in Ax B = 92 tons,    and in a b and a c each = 47 tons.

Case III, is worked entirely from the end view. 100 tons is to, be
distributed on the fore legs, causing no stress in the hack leg. Then
by diagram T, stresses are

in a b and ac each — 51 tons.



WE commence with a few definitions and explanations.
Force and Mass. — Engineers use * gravity' units for these:
he unit offeree being i Ib. and that of mass 32*2 Ibs. (g) or :


mass = —

Velocity is estimated in feet per second. If uniform^ the
listance travelled (s) depends both on rate and time occupied :

US)                 5 = tv   (distance — time x velocity)

shewn graphically at A, Fig. 434. The area B shews similarly the
listance travelled, under variable velocity given by the curve ; the
areas being measured as at Fig. 326, Chapter VIII.

Acceleration (/) is the increase of velocity during each
second. Uniform acceleration is produced by any constant
force, the latter being measured by the increase of momentum
it produces.* Momentum = mass x velocity.

. *. Force producing acceleration = — x /


Uniformly Accelerated Velocity. — A body starting from
rest at o {c, Fig, 434) has its velocity gradually increased by the
amount /during each second /, and the final velocity is 4/ But
the total time is 4. Therefore final velocity,

*-/' ..... (0
and the distance is shewn by the area under the velocity curve

atC>°r:          ,-iJf-i//. ..... (a)

Substituting value of t from (i) we have :


* Newton's second Law.




With an original velocity v the distance is found by adding the
two areas at r>, Fig 434.    (See p. 1099.)




C         sec^-

Uniformly Retarded Velocity is a similar case to D, the
final velocity and total distance being found by subtraction of
areas, as at E, and are vl - v2 and sl - ^respectively.

Collating results, with ^ as original and z>2 as final velocities,

U.A.V. from rest.
	U.A.V. with original velocity.

	^2 =  *'l — //

s = J//»
	>'='/#! +' ^//2
	*=/*!- i//a

V* = 2/S
	V<? = ^2 + 2/V
	^22= »!2— 2/J

Example 4p. — A locomotive and train weighing 100 tons start on
a level, and attain a speed of 60 miles per hour within one minute.
What was the mean pull exerted ?

^        , x     -

From ft)   / - 7 -

and Pull in Ibs. -       =

Io° x


-     . IO266 Ibs.

32 x 15

or 4-6 tons, neglecting friction. ,

Conservation of Momentum; — Two balls, A and B,
Fig. 435, raised simultaneously, are allowed to fall, strike, and
rebound. The duration of shock is called the impact, and it is


Energy Forms.


found that the added momenta of the balls is the same whether
before or after impact, a fact useful in many calculations. In the
case of ordnance the total momentum is divided at explosion.

equally between gun and carriage on the one hand, and the shot
and charge on the other.    Introducing a practical coefficient n,
Highest velocity |  _ (weight of shot and charge) x (muzzle velocity) x IT
of recoil        j   ~~                  (weight of gun and carriage)

the quantities being in Ibs. feet and seconds.

_           ivv .              ,                   f         f        ..      wv      ,

But $ = —~(PP- 473-4)      •'• mean force of recoil = —- and

o                                                                                                                           o>

maximum force = mean force x 2.

Energy is the capacity to do work.—Potential Energy is
latent till some small change occurs to give it actual value: thus
the chemical energy in coal requires a small starting heat, and the
water in a high tank may be released by opening a small valve.
Kinetic Energy or energy due to motion, is always visible so to
speak, except in the case of molecular movement merely.


Energy of position.

/•Elastic Energy.

(Capable of muscular exertion.)
( That due to separation of positively
\ and negatively-electrified bodies,
( as in frictiortal electricity.

fDxie to separate existence of
elements, as in gunpowder




Clock spring ^wound up

bent bow:
Compressed gas:
Nerve Energy:

5, Electrical Energy

6. Chemical Energy

and coal.



Natures Stores of Energy.


I 7. Muscular Energy :

8.  Gas Expansion :

9.  Mechanical Energy:

10.  Electrical Energy :

11.  Heat Energy :

12.  Chemical Energy :

\i3. Radiant Energy:

(When in motion.)

e.g., the wind, heat engines.

As in machines.

{The current in motion, as in
Voltaic and Faradaic elec-

Being molecular motion,
f When combining, on account
\    of affinity of elements.
/The  vibration of the  ether
(    causing light and heat.

The true energy is that only which is available, by reason of a
certain difference of* pressure/ 'head/ or ' potential/ as measured
within fullest attainable limits.


I. Heat Energy:
II. Water Energy :

III.  Wind Energy:

IV.  Coal Energy:

V. Petroleum  or   oil

Energy :
VI. Tidal Energy :

VII. Electrical Energy:
VIII. Food Energy:

f Direct from sun : probably sustained
1    by meteoric impact.

Due to fall from mountains to sea.

/Due to difference of pressure caused
\   by sun's heat.

(Due to chemical condition of separ-
\    ation.


Due to moon's attraction, principally.

'(i.) Due to separation of kind, as in thunder
clouds, and untractable : (2.) Due to very
small difference's of potential in both air
and earth, and valueless for large oper-

(Due to sun's  action  o-n growth  of
(    plants.

All these, excepting VI, are due to the sun's heat, which has
grown coal forests and daily evaporates water. V. is probably
iue to a condensation of the once glowing earth.


* r    > «.*? *'* f il   * tf '^

•/ :".-?;.;



'/f,.r 7* .  *.^r   ;. J

4* ?'i!«?»    i



<**                                          It


»J»   >     |A

r-» |&*^

s***    pt

I, *   ',

I *tf

•> } n

•f    l\         i         t*   rf-*i > ' «   »   «f
^     ,*    '    '    t            ^    ',»*

Prime Movers to Transmitters.              479

A unit of heat will raise i Ib. of water through i° Fahr. when near
39°, its greatest density; and Dr. Joule found by experiment that

One unit of heat = 772-55 foot pounds of mechanical energy.

The number 772 therefore is spoken of as Joule's equivalent (J).
(See Appendix III., p. 930.)

Electrical energy may be estimated in terms of mechanical
energy as follows :

Energy in foot pounds = 737 E Q.

where E = Electro-motive force in volts.          l

and   Q = Quantity in coulombs.   (See App. ///.,/. 926).

Lastly, Chemical energy is measured by its heating effect,
found by careful experiment. Thus, i Ib. of good dry bituminous
coal will give out 14,500 units of heat when completely burnt,
which may be further represented in foot pounds.

Prime Movers are machines which obtain Nature's energy
at first hand for transmission of, or transformation into mechanical
energy. Such are: Heat Engines, Water-Wheels and Turbines,
Windmills, Electric Engines,* and Tidal Motors.

Power is direct or controlled energy, as distinguished from
the free energy of Nature or that, say, of a bullet The term has
been more usually applied to mechanical energy or the mechanical
equivalent of other energies, but is gradually being restricted to
indicate 'rate of doing work,' one horse-power being the unit

Transmitters of Power remove the mechanical energy
of a prime mover to a distance, or change the components and
perhaps the whole form of the energy. The following is a list:—

, Lingerie:                    {

Shaft*    •                      i Lines   of shafting,  with   clutches,

2<     :    m^*                      |    couplings, and bearings.

3.  Spur gearing:              ,    For connecting parallel shafts.

4.  Bevel gearing:                Connecting shafts at various angles.

5.  Worm gearing:               Connecting shafts at right angles.

•4   ti ,    '    *.                      ( Connecting shafts at various angles,

6.  Beltgearmg:            .   j     but chie|y paralleL               *

7.  Rope gearing (cotton);    For high speeds. :

* By voltaic battery or thermopile*


>^' i '^"'r^    *   |   #*^t    '4  ^^^j           |'f   'tl

' /    ///   Hf*>ftt""f9fi/V*    if   <&/*{ fit* ?•**?•*}   t*

4 !i %'
4 f//, i

A              M     <*.$<!

T      / I k |       I I

Simple Machines.


and they can all be placed under two divisions — levers and
inclined planes. There is always a point P where the energy
enters, and a point W where it is removed,"* and the Principle
of Work states that

Work put in at P = work taken out at W
neglecting resistances.    But as work = force x distance,

where d = distance travelled by P, and D that travelled by W.

This is the underlying principle, and our investigations on
machines are for the purpose of finding the comparison of the
distances or speeds at P and W, for by inversion we shall obtain
the relation of the forces W and P. The first is the ratio of
virtual velocities and the second is mechanical advantage. Then,

veLP       force W

-— — = - - ~~
vel. W      force P


= Mech. Adv. -=•*

Mech. Adv.

The Lever is shewn under various forms in Fig. 436.    By
moments :

Pa - WA        and     Mech. Adv. - ? - | ^ ****** Z>

P     • A         p. 763.)

The Wheel  and Axle, Fig. 437, is reckoned similarly,
and its

a         handle

A "" barrel rad.

A train cf gearing in Fig. 438 consists of two pairs of wheels,
a handle, and a barrel. The advantage of the first pair would

be - : of the second pair •—• : and of the wheel and axle —-.    So
A                               Aj                                           A2

the total

Mech. Adv. -=r- = ~ x -A- x »—•
P      A      Aj      A2

* The old letters P and W being retained, are, meant to represent the forces
and also the points of application. Rankine called them effort and resistance
respectively. Note that fractional and other losses are entirely-neglected on
pp. 481-4 and Theoretical mechanical advantage is therefore the result.

(Sup, 954- )

Levers and Equivalents.

which can be easily proved by the levers shewn below.   Generally,
then, for toothed gearing with wheel and axle,

......... .,    ,'•   , .W      followers      handle rad.

Mech. Adv. — = -j-.------ x-------—~~

P        drivers         barrel rad.

the wheels being estimated by teeth, radius, or diameter.

'   l^J^L

Belting is a substitute for toothed gearing, as shewn in the
Diagram, a crossed belt giving the same direction of motion
j&s Ane pair of wheels. N.B.,—If speeds only are to be reckoned,
the wheel andfaxle does not enter into the calculation.

T}ie Brock arid Tackle (Fig. 439).~Neglectirig friction)
the cord has the same tension throughout, and there are (in the
case shewn) six pulls on the weight, each equal to P.

,,   ,   AJ    W      6                           No. of cords

^ .,% .Meet. Adv. ~ =. -    or generally == ——._

v......      '     '    '                 ' W   ' '2

;J   'In-any -movable pulley M,   p- = - because W only rises half

of P, as shewn.    (Seep. 741.)

''W1  *fc    f


4 H     tl         j^t              iti   Jh|

ill **             .*«)»                                 lit

I        '4        i^^^             i,|             |?#               *^ <Uj

*§            a^<$|f  >^|| f* i9^i|f|^| 4*4$

1             "4 t»J ***») j^r|* «ic4| ^| ^yt|(4 ,H|| ir» 1 |s|rf^| t'*M^4 if^*^ * ri| |


M   ^ »*t o |'"*    '*"•*!'     '   4 ', s   '


Screw Cutting*

Ex-ample 42.—Arrange the gearing of a single purchase crab so
that 6olbs. on a 15" handle may raise half a ton from a barrel 10"
dia., neglecting friction. (Eng. Ex., 1891.)

Mech. adv.

W       follower

_ .         x


driver       barrel rad.



1 1 20     follower

_ ___


1 5

_            _ ^

driver        5

follower   __   6*22
driver    ~~~     I

So the pitch line diameters may be 6" and 37'32" for pinion and
wheel respectively, as in Fig. 442. (Seep. 955.)

Example 43.—A shaft A has a spur wheel of 120 teeth, which drives
a pinion B with n teeth. On shaft B is a wheel of 132 teeth driving
a pinion C of 10 teeth. Lastly, on shaft C is a wheel of 48 teeth
driving a pinion D of 8 teeth. A turns at 2 revs, per m. Find speed
of D. (Eng. Ex., 1885.)

The wheels are shewn in Fig. 443.

vel. D  __ followers = 120>jx * 32^*^48 __ 864
vel. A ~" ""drivers           11 x 10 x 8           I

/. D makes 864 x 2 = 1728 revs, per m.

Example 44.—Two men at a crab exert 60 Ibs. each on a 16"
handle. The pinion has 12 teeth, the wheel 72 teeth, and the chain
barrel is 12" diameter. Find the load raised, neglecting friction.
(Eng. Ex., 1888.)

followers x handle = 72 x 16 _  16


Mech. adv. —

W  :

drivers x barrel rod.
16 x P = 16 x 120 =

1920 Ibs.

. 955.)

Example 45.—In a Weston block the diameter of the large sheave
i s 10", and that of the smaller 9". Find the load raised by a pull of
50 Ibs., neglecting friction.

W       2 A       2 x 10     20
Mech. adv. -fr — 7—u == —;— ^ —

JT            _rL — JD                 I                   I

/. W « P x 20 = 1000 Ibs.

(See Appendix 7K, /. 954, without fail.}

Change Wheels in Screw-cutting.—-General principles
-re explained at pp. 147 and 212, it being shewn that:

Kinematics of Machines.


Revolutions of mandrel       _ No. of threads perjnch. on mandrel
Revolutions of leadingscrew ~ No. of threads per iF. oiileadingscSrew

in order to cut a definite pitch.    This may also be stated as

followers at L. S.jsnd  _      pitch L. S.
drivers at M end      ~~   pitch M screw

or the pitches and wheels are in the same ratio^ which ratio, being
found, must be accommodated by a suitable train.

Example 46. — In Plate V., the leading screw being. £" pitch, and
the wheels in the set rising by 5 at a time from 20 to 120 teeth, it is
required to arrange wheels to cut (i) a screw of ro threads per inch,
right-handed, and (2) a screw of L" pitch left-handed.

Pitch ratio


Putting 30 teeth on n (Fig. 135) into 75 on stud (£, Fig. 140): 30
teeth on stud into 90 on L.S., we have,

75 x 9° ^ i§ and the handle at n
must be down.

wheel ratio =

(2)     pitch ratio

30 x 30

*±§- - 1

M"" ~ i =

Putting 45 teeth on L.S. and 60 teeth on n; with any intermediate
on stud (say 60) we have,

wheel ratio - 11 _ 3   and the h^nd1^ at n must
-----------    60     4               btup.

Kinematics (of Machines) is a method of attacking
machine problems devised by Prof. Reuleaux, and anglicised by
Prof. Kennedy. We shall proceed to discuss its principles.

Pairs.—The constraining parts are termed pairs because
they always occur in sets of two. Of these there are higher and
lower pairs. The former connect by points or lines, but the
latter by their whole surfaces.

Three kinds of lower pairs are possible: I. Sliding pairs,
as a piston and cylinder. II. Turning pairs, as a journal or pin.
III. Screw pairs, including all screws and nuts. Complete or
dosed pairs have their motions fully defined: incomplete pairs
require further closure, as at Fig. 444, where gravity is not for
the moment considered.

MUftf**s1»4$t?K     I   'I   A

*  /

*               f

t,        I I*

f      2       t   ' 1

f *)?*•'

*   *


§               | ?     {*.*"$

-   $    4,     .,

Aft                 'I*|

4    ",*,)*         ^^    f   fsf -

t    '*''   ^   **i»,l*.     J

| fe$ i^t?1*

ft*                              fr*f*

**l                an I

*»'•                             J*i» I trt4#J*J»N



*/f rl

fiiiff  ^

i * ' ^     ^  ., ,,,,   ft
I     h'   ' i^ > f i <,* < «t

i *^ „ i, :^r



'   f

m   *

r, ,

Augmentation of Chains.


Flexible links are called tension elements; and fluid con-
nections, as between boiler and engine, or accumulator and
machine, are termed pressure elements, but the latter are always
connected to lower pairs. A pump is kinematically the same as a
ratchet, the valves being equivalent to pawls (see Fig. 450).

Augmentation of Chains is the multiplication of parts,
for convenience or the reduction of friction. Trains of gearing,
and anti-friction rollers (Fig. 451), are examples.

Summing up, mechanism may be divided into simple chains,
formed of rigid or flexible links, which are again united by higher
or lower pairs ; and all chains must be dosed, either by the chain
or by external force.*


Lower ( *• Crank chains :

pairing, t 2, Screw chains :
High   [3- Pulley chains :


4.  Wheel chains:

5.  Cam chains:

1 Sliding and turning and   screw
J      pairs.

Tension and pressure elements.

Uniform motion.

Variable motion.

pairing. ^ 5. Ratchet chains:    Intermittent motion.

A driving and working end are recognised in each of thesey
corresponding to P and W respectively, and the

Velocity Ratio of P and W in Kinematic Chains
will now be investigated graphically. Considering the instan-

Friction closure is one form of force closure.


Curves of Velocity.

taneous motion, in direction only, of the two ends P and W of a
link, each poiat may be supposed, for the instant, to be travelling
in a separate circle, whose radius will be at right angles to the
aforesaid motion, and the two radii will,, unless the directions of
motion are parallel, meet on one side or other of the line P "W.
The meeting point is known as the instantaneous or virtual
"centre, and the ratio of the velocities of P and W will be the
same as that of the radii Jrom the virtual centre. Of course these
may change at every instant, and the centre itself will move along
a path known as the centrode.

Crank and Connecting Rod (Fig. 452).—In'the position
given, W is travelling tangentially, and w D is its virtual radius,
while P is moving towards A, and has a radius P D. D then is the
virtual centre, and at the instant considered, the movements being
along the dotted arcs, pl iv^

vel. P __ D P
vel. W ~" DW

Taking various other positions, we may obtain a series of
virtual centres; and through them draw the centrode E D F, where
E and M are the positions of P when W crosses the line K. E. The
curve passes out to infinity at F and o, reappearing at L and Q, the
direction heing given by the line JP when W is at G and P at H.
This means that P and W have then equal velocities. The
relative velocities being found for any position, their inversion will
give the relation of the forces P and W.

Curve of Velocities.—It is often required to construct a
curve of velocities for one of the points, when the other mows
uniformly. Taking the second diagram in Fig, 452, the triangles
w c A. and w P D are similar, so that

vel. P _ JDJ> _ A c
vel W """* D w "" A w

Assuming "W to rnove uniformly, being provided with a fly-
wheel, A w will represent crank velocity^ while the projection of
P W upon the vertical at c or £ will give AC or &c the piston
velocity, In Fig. 453 the value AC is found and transferred to
the line A w at A E, and this being done for all positions, the ovals
or polar curve may be traced, whose rwttus meter always shews Ps

Time and Distance Bases.


velocity for the given position of crank, while the crank arm itself
gives W's velocity. Taking various positions of P on HJ, and
setting up the corresponding polar radii, the curve of P's velocity
is obtained as H K j, while the ordinates A w, set up dotted on a
base N o of half crank circle circumference, shew crank velocity.


Assuming P's pressure as uniform, the ordinates Im will give
a curve of pressure; and the AE ordinates, being transferred from
the polar curve to the base N o,. will give a curve of tangential
pressures on crank. Notice points Q R and s T, where P and W
have equal velocities, and also points F and w, where P has its
highest velocity, and W its greatest pressure

Time and Distance Bases.—The pfofile of velocity
curve depends on the terms in which we state the base-line
divisions. The curves in Fig. 434 are drawn with a time base
line (equal titties), but the oblique lines at c and D -would be
parabolas if a distance base (equal distances) were used. In
Fig* 453, K j is a distance base, but supposing NO to represent

L I.

Acceleration Curves.



piston travel, N k o would be P's velocity on a time base. The
ordinates at corresponding times are always the same, but the
abscissae vary, and the two cases must be thoroughly grasped by
the student.

Acceleration Curves shew the rate at which the velocity
is changing. Let a point move from A to B, Fig. 454, with
changing velocity, as shewn by the <!urve AC B, AB being a distance
base (here a necessity). Draw any tangent D E F and a normal
E G, drop the perpendicular E H, and turn H G round to line H E,
giving a point in the acceleration curve. Continuing the con-
struction for various points, K L M is obtained, whose ordinates
shew acceleration from A to L, and retardation from L to B.

N.B.—If velocity and distance scales are the same, the ac-
celeration may be measured to the same scale; but, if otherwise,

v = ft. per sec. of velocity to one inch,
d = ft. distance                 to one inch,

a new acceleration scale must be made, being the velocity scale,
stretched or compressed in the ratio -. (See Appendix //.,


p. 863, for proof'; see also p. 674.)   (See pp. 932, 1099, and 1106.)

The Oscillating Lever is examined in Fig. 455. The
virtual radii are drawn : w B a normal to the circumference, and
p B perpendicular to w j. Then :

vel.  P

B  P


or as


For AD being || to j B, the triangles WDA wj B are similar.
Turning A D round to A c, we obtain one point in the polar curve,
found as at Fig. 456, where ADW is right angle. W's velocity
being uniform, the polar radii shew P's velocity. The centrode
curve passes to infinity at K, N, G, and P, the direction of the
dotted lines being at right angles to w p, when the latter is tan-
gential to the crank circle, namely when P and W have uniform

Whitworth's Quick-return Motion (Fig. 457).—
B P being the driver, revolving uniformly, the angular and linear
velocities of W are to be found. Produce p A to c and p B to D,

"I '





a     9           1

£  *

»  %   »•

? I

*     ^

S   *  „»



^           **




Stanhope Levers.

velocity, the radius vector shews P's velocity. The motion of P
is known as pure harmonic, and occurs often in natural science.
Transferring P's velocities to a distance base gives a semicircular
curve, but on a time base forms the curve of sines.

The Beam Engine linkage is shewn in Fig. 460, with
centrodes and polar curves. The lines A p, B w, being at right
angles to the direction of motion of P and W respectively, will, if
produced, give the virtual centre M. Then if B K be || to AP,
the triangles M P w and B K w are similar, and

veLJP     MP
vel. W ^ M w "

B w

the polar curves being completed as before. The centrode curve
only reaches infinity on the side j, when A H, B w are parallel ;
the ends OE meeting at a very great but finite distance. The
polar curves are similar to those of the crank and connecting rod,
P having greater velocity than W at times. When in the form 3,
Fig. 449, the quadric chain has its virtual centres always at infinity,
and therefore P and W have like velocities. (SeeApp. II., p. 863.)

Point paths are often of more importance than forces, but
can always be obtained by drawing the links in successive
positions ; and the mechanical advantage of a complex system is the
product of the advantages of its parts. Taking now the power
transmitters in order,

(I.) Link work is suitable only for short distances, as in
the case of locomotive coupling rods, and is rather a modifier
than a transmitter. We shall take a few further examples.

The Stanhope Levers, Fig. 461, were applied by Lord
Stanhope to his printing press. Two plan views are given : at
first P and W have nearly equal velocities, but when they have
moved to the positions Px and W1? the latter has* no velocity, while
the former has yet the original motion.

P's vel. _ i         j

"""   an




This means that a very great pressure is exerted at W when the
paper and type are in contact. A polar curve for W's velocity
has been drawn in the right-hand diagram, considering P's velocity


Toggle Joint.

constant. D is the virtual centre, and D p, D w the radii; and the
triangle p E B being similar, p B may represent P's constant
velocity, while P E shews that of W. The latter being transferred
to B p, gives points in the curve shewn; reaches infinity in the
direction B/, and nothing in the direction B/r W is then respec-
tively in the positions w and wr

The Toggle Joint has many useful applications, the stone-
breaker and wagon-brake (Fig. 463") being examples. In Fig. 462
the joint is seen to consist of a simple slider-crank chain, o is
the virtual centre, and OP, o w the radii. Producing w p to c,

vel. P      B p

vel. W


and several points, such as F, will form the polar curve B F D,
showing W's velocity, where P's velocity is uniform and repre-
sented by B P. The curve is a semicircle, having A as centre.

Cooke's Mine Ventilator in Fig. 464 is a case of the
quadric chain. Crank and shutter shafts are connected by link
CD, and AB is a fixed though virtual link. Two positions are
shown, the shaded air being drawn in, while the dotted air is
pushed out by the rotation of the drum.

Quick-Return Motion.—See Fig. 457.

Valve Motion for engines needs examination only for
point paths, and will be treated in Chapter X.

Parallel Motions should strictly be termed straight-line
motions^ but are now best known by the first title. Watt's
(Fig. 465) is the simplest. A D and B c being equal, the upward
movement of p will be vertically straight, because D curves to the
left by the same amount as c deviates to the right. This is
extremely near the truth when a is below 20°, but not absolutely
so. Thus :—


-j  (I -COS a)

sin /3 = sin a + — (i - cos 0)

Deviation of P from )         r
vertical                    ~

f              ^^ n\

/   = ~ (COS a - COS /3)

assuming c D to be vertical at central position.




*# -ji ~ ft i


It?   —   4 I

31*1            *C|« i^4|j^i'*«(|

.r*ji ^^4^|>*. j|ii

-          %.  t

J*      "


V ti?"


Feathering Paddle- Wheel.

or z : z1 : : a-^ : a, and the triangles are similar, so that angle a ~
angle /3.    But a is a right angle, being in a semicircle.

/. Angle ft is always a right angle, and
pr is a straight line.

Scott-Russell's motion 5, Fig. 447, merely copies at A D the
truth of the slide c, D A c being always a right angle. A more
convenient form, is the

Grasshopper motion, Fig. 467, where the slide is replaced by
a long link. The gear may be formed (i) with AB = BC = BD
as in Fig. 447, or (2) A B : B c : : B c : B D, the second being used
in grasshopper engines and the first in a steam crane built by
Messrs. R. & W. Hawthorn, where a piston connects directly with
r> to lift the load. The relation of the links in case (i) may be
found graphically : produce points D, B, c, to the respective
positions i, 2, 3, on the base line i, 3,: with centre 2 strike arcs
i, 4, and 3, 5 : join 4, 5, and draw 5, 6, at right angles to 5, 4.
Then 6 produced gives point A, and length of A B ; for 5, 2 is a
mean proportional between 6, 2 and 2, 4.

The Feathering Paddle-Wheel is shewn in Fig. 468.
If the vessel move to the right with a velocity vs while the wheel
rim has a linear velocity of v/t the floats should enter and leave
the water in the directions vr if they are to meet the water with-
out shock, for vr is the relative velocity of float to water, found
by completing the parallelogram. The controlling mechanism
is obtained by quadric chain H G K E where H G is the fixed link.

Stresses iri Linkwork Members may be ascertained
from the principles in Fig. 423 et seq., the structure being
balanced by known external forces.

The Work Done at any point of a machine is obtained as
at Fig. 325. Taking the case of harmonic motion for donkey
pump, let total piston pressure P be uniform during stroke d':
then P*/' = work done at P and is shewn by diagram in Fig. 469.
Setting out trie pressure-curve for W, on a base ?r R, as explained
in Fig. 453, tne mean of the ordinates will be found to be

•636 P, and as                                      ,       .

work put in = work taken out

P X   2R » '636 P x TrR

A rrangement of Shop Shafting.                 501

which are equal, or no work is either lost or gained  in trans-
mission, if friction be neglected.

jq. 4 69.

(2.) Shafting is used extensively for power distribution in
workshops, being combined with belting and toothed gearing.
Fig. 470 is the plan of a small shop as usually arranged. The

j *   yi ^fc *$       toA|$ I*!

|t      t'f'W^                  f

I'^"v"**'*»* ^*"*»¥ *

*•   j,'    4«,   ,f •*>*',. V

J   ,*t      f*   ;    '»  '   >»',•      -.    t

m      {    ! ' ^r-

i    •< / | ><' #i j    ;».iH*| *i*  * >f

5 *    /    ;• |    //#    *'««'    '*"K(4
*       •,  f * A  4 :*»»> /^i i"   ;

f      *      - » I  i t^             » ;  ' K ,    %    '

i; »•


.^_v_.......... ,.,,,.:. »



Keys and Bearings.

'claw ' clutch, c is fixed by key to the right-hand shaft, and B
slides on a pair of feather keys D in the left-hand shaft, so that
the claws at A may be locked or unlocked. The clutch strap E
encircles the clutch B, and is further grasped by the fork lever:
this gives a sufficiency of wearing surface between the rotating
clutch and stationary lever. The difficulty of entering the jaws is
met by the adoption of friction clutches. (Seepp. 569-70.)

Two shafts slightly out of line but mutually parallel may be
united by the Oldham coupling Fig. 474. A middle plate c, having
cross strips, unites with grooves in the flanges A and B, and the
velocity isl transmitted unimpaired. If the shafts are mutually
inclined, the Hooke's or Universal Joint, A, Fig. 475, must be
employed, and if considerably out of line though parallel, B must
be used. A transmits the velocity unevenly, but the double
arrangement B -rights this difficulty. Fig. 476 was adopted for
many years at a northern establishment: E is the engine, and
u j are universal joints, while the three shafts represent three
separate shops. (See Appendices L and /Z, pp. 763 and 866.)
^ Keys were examined in Figs. 374-5. The sunk key is best,
but the flat key is more often used in shop shafting. Cone Keys
(Fig. 473) are made from a hollow cone, turned and afterwards
divided : they give a very perfect grip.

Keys should have a taper in depth from front to rear, and a
gib-head adopted as in Fig. 477, if there are no means of otherwise
releasing the key. Although some workmen fit keys at top and
bottom only, they should no doubt fit accurately both at top and
sides. Shrinking boss on shaft gives very great security. Keys
are sometimes forged on the shaft. (See p. 423.)

Feather or sliding keys can be fastened either to boss or shaft
as most convenient. See A and B, Fig. 478.

Bearings are strictly gun-metal supports termed bushes, but
the supporting brackets take various forms. Fig. 479 is a
common hanger, Fig. 480 a wall box, and Fig. 482 a wall bracket.
The last two have bearing and bracket separate to allow of adjust-
ment. Fig. 481 shews a special hanger, having a long cast-iron
bearing lying in a spherical seat which adjusts itself automatically
to the shaft deviation. Permanent vertical adjustment is obtained
by screw and nut

' yHf'!"<&"*
|                               1    4*1

/v       rrr.

/    ,f    ^       *    4       I        0 ,...**•

sjt^wsa^ww^iww^s         **V                            f^          .          ^

•  I Tr*4' "il /   , M

\. \     :!. •/


r»t           ^>p (^t                                1 4


(       J   ^

1*1   f        i     S  f   *    ?   i,*   f **|^fe
f * S i /•   . f jr ,4   *f ^,t

f fK| %i **s§i


IVif  tiff              ffi   l*|/          Jf

M ,               t# 4    |

|i#| *'|   HI ,                          41**                                                $sr                      f^,

*^ |1^ ,           *l |4 fe   f^                          <4

Iff |

by                                *

I'll   fc»  ^             f» llw               II  '

ir v I f *


.508                            Square Shafts.

w being exerted through 2?r feet at every revolution:

rT ^


X 27T N

s 7T ^8        27T N


33000       12x16     33000     320^10


= ^320810

_    __
tJ   p






Example 47.—A shaft transmits 20 H.P. at 160 revs. Find (i)
how many H.P. it will transmit at 250 revs., and (2) dia. to transmit
40 H. P. at 250 revs, withy at 2 tons per sq. in. for stiffness.



H.P. ex

/.    20 oc loo   and    H, P. req. a 250
100 : 20 : : 250 : H. P.   and H. P. = 50

d= 68'44




Ex-ample 48.— Compare the weight of shafting in a twin with that
iri a single screw ship, neglecting couplings: the H.P. in each being
the same and the speed of each twin being 25 °/e above that of the
single screw. (Hons. Mach, Constr, Ex., 1886.)

d oc


"H   P


... a i for single shaft ... oc      73 for each twin shaft.

- « i for single shaft ... «


Square Shafts are often adopted in travelling cranes. In
Fig. 490, B Is the longitudinal and A the CTOSS girder of a crane,
the power being given from shaft D through mitre gear to F, and
by spur gear to G. As the carriage moves along B, the tumbler
bearings are turned through a right angle, and are only off the
shaft during the passage of the rnitre wheels, the bracket at B
being shaped to serve as a tappet.

Spur Gearing.


Long screws sometimes serve as shafts, as in large planing
machines with travelling tool, and a linear advance of the screw
may produce rotation if sufficiently large in pitch, as in Fig. 491.

f. 49O.

(3.) Spttr Gearirtg transmits power between parallel shafts
only. Spur wheels are the equivalent of friction discs, having
teeth provided to avoid slipping with heavy loads. The teeth
are formed partly above,and partly below the disc outline, the
latter becoming virtual only, and being then termed the pitch
line. Thus,

Cycloidal Curves.

Pitch Circle, Li|ie, or Surface of a spur wheel
represents the contour of the ideal disc or straight-edge iuh£c2* *"^jt
transmit the same motion.

To transmit perfectly uniform motion the teeth rtiu& £
specially formed, and all teeth in gear at once must contr'i
to the perfection of the motion. To fulfil these condltioti
normals to all surfaces of contact must pass through the m^
point of the pitch lines (Fig. 492), and this is obtained whet>
tooth b c, on A, is the envelope of the relative positions of the o *t*
tooth on B (Fig. 493) when the discs are rolled together,
teeth are actually drawn, however, in a somewhat
manner. (See Appendix ///., p. 926.)

Cycloidal Curves.—A cycloid may be traced by   a
on the rim of a disc which rolls along a straight edge,
epi-cycloid when the disc  rolls  upon a circular arc  (Fig.
A  hypo-cycloid is   similarly  traced  within an  annular   d.ls^
at   Fig.  495,,  noting that  when  the rolling  disc   is     half     tJ
diameter of the annulus a straight line is obtained,    as
dotted;   a   fact  which  has  produced  White's   parallel
(Fig. 496).

Rolling Circle.—The above curves will serve for
teeth, if the same rolling circle be adopted for parts that
in contact, the tooth point being formed by epi-cycloids
the root by hypo-cycloids. Taking the wheels A and B, Fig.
a rolling circle is first to be chosen as governed by the
curves: thus, if the circle be half the pitch diameter, radial
are formed, as at c; if larger, the root will be undercut as at: \
and E is drawn with a circle of J-pitch diameter. The la 11
is reasonable, as giving strength, while yet avoiding ot>llc|
pressure on bearings. 'Adopting then the rolling circles sto^m
F may roll the root of B and the point of A, because these arc*
engage, but G will serve for root of A and point of B. Wlfcoxi ,
the wheels of a train are to work together interchangeably, t
same rolling circle must be used throughout If the toe
pressure is always in one9direction, as in Fig. 499, a large roll!
circle may be adopted for the acting surfaces and a small o:n.€i j
the back surfaces, thus giving great root strength without o"t>lic|


|                ||         % >*«|

W       |f W*'$f jf */(**»            ^f|

if   '^fs|   v*W*t,£                  *i   *r   wit    '{ 4   4|fi||                               ft   }f% I

»•/.'**   ,**    A**    ^|    JW   *   |M«                          J^    ?i#f»     Jtffr                                      4    §

Jf   1   i,    *^  ^'^   J6**!*/1   ^/   ,*#ff    »•'*              ^||        S **J>|f *   iJ^Wlt   J^vrwl

|f     |    t^^«|    |»i#l*   **| * I

rf *>|^**I|PH1 #|t i   4ft   $|>«

>*       Ifp        >                 *|l«r^f|



\   //;**'   ^fi *

If! *     J ' -

.*«.' *    *      J.U   ' ?

/n;   i-   ,i     i

i* -   <,>           N ^-«


#"«                   ^iim |*«w

«|tf.4 wt If I** *f»**t |t Jf f a**?*!

<*|f <i|%*f*4                     Ji * H|*»   »^f '<tf   ' f^ffft          jot

f%f 4t|         t   | *****   «w |A*|**f**»  IN         . |/      '4                ,|

f|0j|,   Ji|| wj«
Sjt                     j*M



*'| I



t i

fS                              'MM               l«    *#

*   f'i*NPWt  »#« %?   ^f^1*!!          I%f|ii4

/ v


,| If

•f II

^ ft

tt>'*   |

t |«ti
y ty*t

*4M ,f

'^,     ft        *

f It

«% Jt

.f   $f

/J   '!&»#

«   \       f

-         |f

., t  - i i *.*>>* it*



^^^fr^    r|             fr*idy''

*l|t   MM

«l                    j^ii^   m   fij   in

J^t                                                      |/*f*                      Kit                     ftttf

f - §  j?f ']*4,*ff **f* t'fr                                                                            |tf

li^#1!ift                           ^*,#lfl/ rfyj            #*.l  Iff $*                             <ftf


*/  JF i    /**   /


>*»*- < '^    ^


i H

*'/?' ***<•(

i'rtf >»/'/

...... n ~1*r"


ii j ^fiipi *^s* ^** "*?"*>   #ff

i#. f 4M<I'4 ««^l **>^ li|,i*f  J || I****  *»l
^»{| *w i*«l   i| ^.^'.|   f i    » , /",    ,|   ^   vt i *,il

-»'t v*                                                          f<> ^it

Bevel Gearing.


nre machine-moulded by two half-patterns, and work smoothly if
well formed; being said to be stronger than ordinary teeth,
•u-hich is doubtful.

(4.) Bevel Gearing connects shafts whose directions meet
.at any angle. Their ideal form is that of the frustra of cones,
.as A and B, Fig. 512, having a common vertex, as c. The pitch
-diameters are measured at d d


Two shafts A and B, Fig. 513, are to be connected so that
their revolutions shall be as 2:1. Assume any convenient
-diameter c 3> and draw c K and i> M |[ to A L. Taking E F « 2 c D,
draw E H and FG f| to BL. Through G draw GH at right angles
to B L, and G K at right angles to AL : then join H, G, and K to L.


/.V ^*->              *Ji*

***& i^r.«  «H
)f           ^|                 ^»

ftyfr'jS  j>ff %#*f|f   V**V

«-• -     »**'•   f iff t      j,^», fff'ttr*r*

V*        /     fl^'        « S     .                                           **                              *W     ,f     *"          >t"      »       r-      ^|      .»^tH

I       '-114   ^'*Mf           >    r,   ;    i   t«M

'    ,    *    I     , ;< M f | I *      it;              §

. (    *   4 i i /" «r i    I?***)'     ,     t     i     >> |         , ** 4   * ^ |



1 'ff, n /*/*«*

»..» if   ftfv


Screw Gearing.


gear, with so many threads to the worm that it becomes a wheel.
Fig. 516 shews its application in a Multiple Drill where A A are
tfce drivers, and B B follow on the drill spindles r> D. The wheels
are here equal, and the teeth are inclined at 45° to the axis.

Epicyclic Wheel Trains, like worm gear, produce a high
ratio with few parts. Kinematically they are ordinary trains
where one wheel is the fixed link.

Case L — Fig. 517. Let A and L be in gearv with AL fixed.
If A make a minus rev. with relation to A L, L will have made

- plus revs., because — is the ratio of the train.    Next fix A and

L                                        L

put L out of gear. If now arm A L make one plus rotation two
things have happened : A has made one minus rev. relatively to
A L, and L has made one plus rev. relatively to A. Finally, put
A and L in gear, and give A L one plus rotation . L receives two

one plus rev* due to its connection with A L, and - plus



revs, due to the relative minus turn of A

A, and        «

I/s revs, «

both relatively to


+ =•

*r          V {

,***-%'4*  ^   I**   4

r  * %



1    t      /TV**

vi   t  V.   /


Reverted Trains.                           523                     *

relatively to A.    We  may vary the experiment   by carrying A                       ^

round L, but so that A does not revolve; then the relative posi-
tions will still be the same, as shewn by a comparison of the                      , i
figures, and L will again make two revolutions while A is carried
once round it.                                                                                                     •

Watt's sun and planet gear, Fig. 521, is a practical example
A slight deviation from the rigid vertical occurs at c and d, but
the total result remains; s makes two revs, for one rev. of the
crank.                                                                                                                                       ! j

Case IV.—A Reverted Train is where A and L turn on the                         \

same axis.    In Fig. 522, A is fixed and L reverted, while abc                      l'j

shews the train in direct order.

.   .   a x c

1 he tram ratio is -— and
b x/

n                  a*c

/ s revs. = i - -—-

If A  and  L are nearly equal, we may obtain a very slow                           fyj

relative rotation, as in Fowler's first coiling gear, Fig. 523.    Stud                           'i

D supports the drum and gear, A is the fixed wheel, and a dif-                             \

ference of about one tooth in 40 between A and L causes the                           ^1

latter to turn very slowly, rotating the cam  E, and raising or                           Jfji

lowering the coiling lever and guide pullies as required.                                              |l

Fig. 524 has an annular wheel, but is otherwise like Case II.                             „
Opening out the train, it is found that while /'s revs, are minus,
those of L are plus, so                          .
L's revs. = i 4- _~

Its application is shewn to a ship's capstan ; and

TT , ~    .       L's revs,     lever arm
Vel. Ratio = -—— x  -      ,

i          barrel frad.

D being inserted for steadim en t.       s

MoorJs Pulley Block, Fig. 525^15 a ^evirted train with annular
wheels.     Referring to the  lower  diagrams, the  train  ratio   is

~2-~ and a miniis rotation is induced, in / or L by the relative

motion of # or A*                               A x C

.*.. L's revs. » r ~ ~—^
.                                   B x'L


^ ^*^%^.   ^w*.      ^ -we
•*»-.      „**
«                   «=s^*^».                 -jf.                          *5.«j»-'»«*     -^^*^-^


526                        Moore's Pulley Block.

If A and L are nearly equal, we have a high velocity ratio.
In the block, the eccentric G, corresponding to crank ef, is rotated
by hand chain round H, so that A and L are turned oppositely,
each by half their relative motion, and w's rise is due to this.

P's distance = 2 ?r R

,TTi    ,.               2 TT r x L's revs.

W s distance =------------------


i xr i  ^   -        p's  dist-           2R

and Vel. Ratio =

W's dist.     r x L's revs.
In the example BC has 14, A 15, and L 16 teeth.    If R = r
Vel. Ratio =        ?•—• = 32 : i

1        14 X 16

Another reverted train is obtained by bevel wheels, as in
Fig. 526, being applied as driving gear to traction engines and
tricycles. B is the arm, and A, L the first and last wheels respec-
tively. When the front road wheel is steered ahead, A, B, and L
are practically locked, and the two hind road wheels move with
equal velocities; but if the front wheel be steered, say, to the left,
A becomes fixed and L revolves at double speed, thus steering
the engine in a much smaller curve. Fig. 527 shews a detailed
section through the hind axle.

Fig. 528 is a disguised form of sun and planet motion, where
L is annular and the slider-crank chain is employed. Considering
A fixed, as in Fig. 520,

T>                              A

L s revs. = i - -=-

If A an<J L are nearly equal, a slow movement of L is obtained,
as in Fowler's second coiling gear, Fig. 529. Eccentric B serves
as crank, and D as connecting rod; A and L have the same
meaning as in Fig. 528, and the cam and lever are as previously
described. (See p. 1108.)

(6.) Belt Gearing has the disadvantage of slip, but is
practically .noiseless, and will transmit power a considerable
distance (say 30 ft.) without intermediate support.

I            |»*

IK#   1*                4* ** f*

f               0                 '<%

^  y             * f f


^tM^f    *i^l
^.ft'?*|   flv/^?ia   c^*4/«t

f ^, ^ ,fi|/ | «/;,   f         t4 * ^JNP'W^    *f rl* ^' *•






Tension of Belts.

make a constant angle with the radii, and if the angles cr are taken
small, the construction is fairly correct. But greater accuracy is
secured by using the equation to the curve,


where e = 2718, the base of Napieran logarithms.

More usefully the formula becomes


/     (See Appendix //.,
•«43 M-           /867.)

Then, the  log.  being known, the corresponding number is
found from  a table, and  the  formula used  for  any  value   of

-, even beyond 360°.


Leather belting" on iron pullies .........
	"\ to *A
	i "5 if oily

• Wire rope or^iron pullies ...............

'Wire rope 0n leather-bottomed pullies Hemp rope on jbron pullies    ....... .....
	•25 '28 to *i8
	) not accounting >    for wedge \    action


Driving Pull of Belts.



 -      L°g
 £  :   L°g
		Tn tn

l£    I    '09691

1^  ;  '17609

rf    <243°3

2     -30103

2|     -35218

2i   '39794
	j  43

2I : "43933

3  : -47712

3i ! '5^88

Driving Pull and H. P.—If two weights are slung over a
pulley, as in Fig. 531, the pull on the rim of the latter will be due

to their difference, wl - w, and as this is the same case as a
driving belt,            Driving pull = Tn - /„

and H. P. transmitted =


But V = 27rRN

.-, H.P.


Strength of Belting, allowing for the joint, may be taken.

so that

/"^(safe) •« 320 Ibs. per sq. in.

and the thickness varies from -^" to f" in single-ply belts.    The
width must be made sufficient to meet Tn.

530                  Centrifugal Tension in Belts.

Example 53.—A leather belt is to transmit 2 H.P. from a pulley
12" diameter on a shaft making 160 revs, per m. Find (i) the
tensions, when the belt embraces half the pulley rim, and yu = -3 : (2)
the belt width when the leather is £" thick.

(1)  Log. Is = -4343 x -3 x 22 = -40905       .'. I: = 1*.

/n                                 7               .                IK         I

TT ^      (Tn-4)27rRN        , _     ^        2x33000x7             .,

H.P. = ^-2---------------andTn~4 =         -          '    — 131 Ibs.

33000                               2 x 22 x-5x160       °

There are two values of Tn, viz., (44- 131) and (2-5 4)

.-. 2*5 4 = 4 4- 131 ;      4 = 87-3 Ibs.       and Tn — 218*3 Ibs.

(2)  w" x -25 x 320= Tn = 218*3       /. w" — 2.$" (Seep, mo.)

Tension in Belt due to Centrifugal Force may be
examined similarly to the fly wheel at Fig 353. The weight of a
cubic inch of leather (w) is '0358 Ibs., and the stress per square
inch becomes

I 2 U)1)   11


.-. Total tension on tight side = Tn 4-  ^w* *>"

which is the total force the belt must resist at high speeds.

Creep, Slip, and Speed.—As the belt tension changes
from Tn to /n, a small retrograde movement or creep occurs
due to release of tension, causing the follower to revolve at a
slightly decreased rate. The result is known as slip, and repre-
sents a loss t>f about 2 per cent. The speed of belting should not
exceed 3000 to 4000 feet per m.

Length of Belt (Figs. 532 and 533).—The length between
centres c should not be less than 6 times D if much power is
transmitted, though much less is used with light pressures. It
may be as much as 30 feet. Horizontal belts give better
results than vertical ones, and some inclination should always be
given upright belts if possible. Taking the oj>en belt, Fig. 532,

and Total length of belt = 2/t + /2 + /s


>?*>>%>*$*    *

rftff ^fn


{ *   *   * r-*         !/'    '*   -/   }*   i»jrf  * a  |*4 |   />.«{*   /f J|ff!' I  #


•        '   ,   ,    »     \ { t* * *    t * it i •   f fp j J 4t ^ ;•; f  *^|

*      **     |          *' ,  ,'     '   s   (     '*  ,  *   '-   4     (  '  *            ^ |'*|^i     * | »

11 •*

Counter shafting and Putties.

movement is leftward, and the final position is that at B. The
radius of curvature should be three to five times the pulley

Countershafting and Speed Cones.—Fig. 539 shews
how a shop machine M may be driven so as to be started and
stopped without affecting the main shaft revolutions or removing
the speed cone belt c c. B is the main shaft and A the counter-
shaft, the latter having fast and loose pullies L and F. The fork/
on the striking bar s then grasps the advancing'side of the belt,
and is moved to right or left by pulling the handles D, which act
on the belt crank L.

Quick return is obtained by the belting at Fig. 538. An open
strap turns the advancing, and a crossed strap the returning pulley,
and in each case there is a narrow fast pulley and a broad loose
pulley. The fork is shifted automatically at either end of stroke,
and the machine stopped by placing both belts in position shewn,
from the handle H. The total width of pullies may be reduced
to four times belt width by the arrangement shewn below, where
two striking bars are employed with which the black tappets
only engage at certain times. Many belt examples will be found
in Part I.

Problems in Belt Driving.—The more difficult cases are
solved in Fig. 540, and will be understood if it be remembered
that the advancing side of the belt must lie at right angles to the
shaft, while the retreating side may make any deviation.

Pullies for Belt Driving are usually split, for convenience
in fixing. Fig. 541 shews the construction of a cast iron, and
Fig. 542 of a wrought iron pulley. The former should have
curved arms if more than 12" diameter (see p. 67), and the latter
is adopted for lightness with high speeds or large pullies. Fig. 543
shews a section through a pair of fast and loose countershaft
pullies, which need not be split.

(7.) Cotton-Rope Gearing is much in favour for spinning
and weaving mills, and has been successfully applied to travelling
cranes and dynamo driving. For mills, the flywheel rim has the
section shewn in Fig. 544, and the ropes lie in wedge grooves.

With a flat pulley thie resistance to slip would be P/i, but in
the grooved pulley shewn the resistance is 2 R/u, there being two

Cotton- Rop£ Gearing.


friction surfaces.    The grip is greater in the second than the first

case in the ratio -~
22^° = 2-6131, and jj.

From the force diagram, -~ = cosec.
•28 x 2*6131 = 732, which should

be substituted for yu in the tension formula already given.

Messrs. Jno. Musgrave and Sons, of Bolton, have fitted up a
large number of mills with cotton-rope driving, and the following
remarks afcd tables are the result of their experience as given in

' j|f*f<^           *'P   tf     III     |*"S

*  l-^l     ***   ^'*   *^*t   tl^ff   t   If

«« # |, t /rt **j$ 5f        i


% ^; ^

'" /

,, ;„ .i .1. ,..<„„,„.       /       -»jU« *•'

^t*.,,.,,.,i*mm,>i"'>'"«" i      '>•         If                 f

^^         if,        .f     >'        f





(Messrs. MUSGRAVE.)


Dia. of rope.
	Area of circle.
	Weight per foot.
	3*1 t
 ^,5 *
	Centrifugal stress
	Effective tension
	H.P. transmitted.
	Centres pulley grooves.
	Dia. of smallest pulley.

	sq. ins. •1963
	Ibs. -08 1
	Ibs. 47
	Ibs. 16
	Ibs. 31




*      -7854

1 |   i    I ' 2 2 7 2

i£ !  1-7671

if     2-4053
		1 'OO

2   i 3*1416
	7 no

The centrifugal stress is weight per foot x z/2 -r £-, and the
fourth and sixth columns assume that /n = '2 Tn which gives Tn :
Vn : : 5 : i.    From the tension formula,

Log. 5 or 69897 - '4343  x (/* x 2*613)  x JSJ112L£?



i|7/ is the usual diameter for main rope.

Fig. 545 shews a spinning-mill driven by cotton ropes, the
power being-given to five floors by separate sets of ropes, a good
arrangement in case of breakdown. The slack side being upper-
most gives a large arc of contact. Fig. 546 shews a travelling
crane. The rope is endless, passing round the pullies D, H, A, G,
F, and E in succession, and kept taut by the weight at H. Worm
gear is used in taking off the power, at E for travelling, at B for
lifting and lowering, and at c for cross-traversing; and either rope
is put in gear by the press pullies a, fr, actuated by hand levers P, Q
E is reversed by friction gear worked from handle L.

JVWUL driven,   by  Cottars floras,     Jfliq.


Wire-Rope Gearing.


It should be noted that horse-power, depending upon pressure
x speed, may be obtained either by a large value of the one or
other quantity. Thus cotton-rope driving depends upon a low
pressure and high speed, but high-pressure driving will now be

(8.) Wire-Rope Gearing, introduced by Him in 1851, and
called by him ' telo-dynamic transmission,' has since been used in
many long-distance cases, for example :

1.  From turbines to distant mills.

2.  For steam ploughing.

3.  In collieries: both for hauling and raising.

4.  For travelling cranes.

5.  Funicular railways and cable tramways.

6.  Boat towing on canals.

The rope is of steel wire, with hemp or steel core, and six
strands of from 7 to 12 wires each. The wear is more uniform
if the strands twist in the same direction as the rope, as in Lang's
patent. The following table refers to the latter ropes :

Circurnf. of rope.
	Dia. of circle.
	Breaking Stress.
		Dia. of smallest sheave.
		Hemp Core.
	Steel Core.
		No. of strands.
	Wires in each strand.




	15 TV



	,     12



The wire core does not affect the safety of the rope in bending:
round pullies.

o o

54O                        Pttllies for Wire Rope.

Pullies for "Wire Rope.—The section is shewn in Fig. 547,
having a groove filled with leather on edge which is afterwards
turned: p then is '25. Fowler's clip pulley, Fig. 548, has its rim
divided into a series of toggles, the mere pull of the rope causing
great grip, as shewn at A. E is a huge screw on the pulley rim
which permits adjustment, after which the bolts are re-inserted.
The clip pulley has enabled wire rope to be applied in many cases
hitherto unsolved. Fig. 549 shews a guide pulley.

At Fig. 550 a turbine (or horizontal water wheel) Tb drives a
distant workshop. A B is termed a relay, which should not
exceed 500 feet, and c c are guide pullies. Fig. 551 shews two
methods of steam ploughing : (r) is the ' direct' system, engaging
two engines which wind up the rope alternately, and advance
along the headland between bouts ; (n) is the * roundabout'
system, where a portable engine A drives a windlass B in either
direction as required, c, D, are self-acting anchors, which resist
the pull of the rope; and as the slack-rope anchor automatically,
winds itself in the direction of the claw anchor F, the tight-rope
anchor is meanwhile fixed. G is a rope porter. Fig. 552 serves
to explain underground haulage. An endless rope is used at (i),
being crossed at j to obtain a greater grip on the clip pulley H,
and tightened at E with a heavy weight, (n) employs a pair of
winding drums c, as in the case of steam ploughing. The haulier
attaches his wagon by scissors grip at A. The up and down rails
are omitted for clearness.

Fig. 553 represents the lifting gear at a'pit-head. The cages
move in opposite directions, and while one drum is winding
the other pays out, a brake being attached to each. When the
mine is very deep, the conical drum, Fig. 5530, is advisably
employed. It is on the fusee principle. When the cage is near
the bottom the load is greatest, due to rope weight, and the
drum radius is decreased, so that an approximately even turning
moment is required throughout the lift. Overwinding has con-
stituted a serious danger, and may be avoided either by
automatic reversing gear on the engine, or the detaching hook
in Fig. 556 (Walker's). The mouth of the hook is usually
closed by the ring A, but if the engine be over-run the hook
attempts to pass through the ring B, in the beam c above the


Stresses in Wire Rope.



pit-head; and A is thereby caught, being slipped relatively down-
ward. The jaws then open or catch on B, as at D.

Fig. 554 shews Fowler's travelling crane driven by wire rope
round clip pullies. A is the rope arrangement, and the power is
distributed for travelling at c c, cross traversing at r>, and lifting at
E F. The last is accomplished by the rotation of screw F, which
shortens the lifting chain attached to nut E. The arrangement is
suitable for very heavy cranes.

Cable tramways are useful for bad inclines. An endless rope
travels in a conduit A, Fig. 555, and the car carries the gripping
lever B, which, when moved to the vertical, raises the rollers c c,
and brings the jaws D D together. Some jerk is, of course,

Fig. 557 is a towing arrangement adopted on some German
canals. A rope is anchored on the canal bottom, and the
tug winds itself along by the engine-driven clip pulley. The
rope serves as a rail, and with the pulley forms a kinematic

In wire-rope transmission the tension ratio is usually 2 : i and
the speed 3000 to 6000 feet per m. The stresses in the rope are
due to:

(1)  Weight of rope and the form of hanging curve.

(2)  Bending of rope round pulley.

(3)  Centrifugal force.

(i) In Fig. 558 the catenaries may be considered as parabolas
for all practical purposes. Then the tangent T A being drawn, by
bisecting c D at A, the force diagram will give the value of T, in
terms of W the weight of rope between the pullies, and B the
pressure on the bearing. The weight of wire rope per foot =
(1*34 x d*} Ibs. (2) Taking the general bending formulae,

El      J     ^
Bm==_==/      /Z

p      y

and I = Zy
where p = radius of pulley, andj' that of the rope-wire:

= 30,000,000 —.




(3) has been already treated for belt and cotton rope. The
safe strength of the rope must meet the combined stress (i) 4-
(2) + (3), but the driving tensions Tn and /n caused by T will
both be decreased by the stresses (2) and (3).

Two shackles are shewn at Fig. 559. At A the wires are bent
back and soldered, giving a joint equal to the rope strength; but
B is wrapped round a wrought-irori eye and then spliced, the joint
having but 50 or 60 °/0 of the rope strength.

(9.) Pitch-chain Gearing serves the purpose of belting
where positive driving is required or considerable pressures
are to be transmitted. If high speeds are employed, the gear
should be exceptionally well made. Much power is lost in
friction, and the journals must be adjustable to take up stretch
or wear.

Fig. 560 shews three forms of chain. At A the teeth bear on
solid inner links, but at B and c they engage with the pins, and
the smaller pitch obtained gives more regular driving. There are

Pitch-Chain Gearing.                         545

two sets of friction surfaces; the teeth on the pin and the pin on
the inner links. B decreases the friction of the former surfaces by
the introduction of rollers, and at c the latter surfaces are enlarged
by riveting a ferrule to the inner links. The pins should be
shouldered, so that the links may work clear at the sides; and
the teeth are involute curves having the arc of pin centres as base

D is a road roller supplied with pitch chain, and Fig. 569 an
electric car driven by chain from a dynamo. Cycle driving is a
well-known application. (Seef.nix.)

(ii.) Compressed Air is of great advantage as a long-
distance transmitter, and as such has been used for motors in
mines, for tunnelling machines, and for distribution from central
stations in towns.    In mines and workshops, the exhaust serves
^ also for ventilation.   To be effective the compressors should be
f on a somewhat large scale, and the arrangement is shewn in Fig.
561, where steam in cylinders A A is used to compress air in the


Compressed Air,


cylinders B and c, which is conveyed thence to a storage receiver
for distribution by main to the various motors.

When work is done on a gas, the temperature is raised, by
reason of the conversion of that work into heat; and again, when
9, gas does work its temperature is lowered, for a reverse reason.

The changes are shewn in Fig. 562, Draw co-ordinates o F for
pressure and o J for volumes; then let the piston commence with
a ^cylinder volume i, opposite A, and atmospheric pressure 15 Ibs.
at i A, the temperature being 60° F. A c is a hyperbola or
isothermal (' at constant temperatureJ) of 60°, while E B and D K

Losses in Cooling.                            547

are parallel hyperbolas at 320° and —201° respectively. In com-
pressing the air without subtracting heat, its temperature rises to
320°, and the pressure curve is the adiabatic (''no heat passing
through') from A to B, the volume being now reduced to '5 with
pressure 45 Ibs. • Suppose the temperature next to lower to 60°
during transit to motor, pressure remaining constant, which is
practically true, the volume will decrease from B to c, viz., to "32.
Next let the air expand behind the motor piston, without adding
heat, and its pressure will fall to 15 Ibs., while its volume becomes
•74, and the expansion curve will be the adiabatic c D, the final
temperature of which is - 201°. The area A B F G shews the work
given to the gas, c D G F that restored to the motor, and the loss
due to cooling is the area A BCD, being here about 27°/0- In
practice the curves would more nearly approach the thick dotted
lines, but there are losses in steam cylinder, main, and motor,
which may reduce the efficiency to 30% instead of the 73°/0
shewn. (See also fp. 773 and 881.)

Formerly simple steam-engines were employed as compressors,
but these are now replaced by compound engines; much im-
provement too, has been made in the methods of cooling. It
being granted that isothermal compression is the ideal condition,
the old water-jacketing proved inefficient, as removing the heat
after adiabatic compression had been permitted. Water pistons
were little better, being cumbrous and slow; while water spraying
in the air cylinders both spoilt the cylinders and gave but a slight
further advantage, for the time was too short for the heat
to be taken up. The greatest improvement was made by the
introduction of two-stage compression, or the performing of the
work in two cylinders, with an intermediate cooler. It can easily
be shewn that if a succession of such cylinders and intermediate
coolers be used, the compression may be truly isothermal, thus
gaining a large portion, but not all, of the lost work area, for loss
in the motor may still occur. The blocking of the ports with ice
or snow on account of the low "temperature of the motor exhaust
caused trouble ..Which an attempt to overcome was made in 1887,
by re-heating the air entering the motor, but .th<^ economic results
proved such a surprise, (:hat the method has ever since been
followed, with increasing success; and it is now clearly under-


548                            Compressed Air.

stood that the employment of fuel in a re-heater is attended with
some six times the economy of the same fuel used in a steam
boiler furnace. Figs. 563-4 shew at B an electrical re-heater
formed of resistance coils in the circuit of the dynamo A, and at
G a stove re-heater through which the air pipe passes. Still
another saving is obtained by air injection. As heat is nothing
but a form of work, it may be made to do work as soon as
generated, instead of being allowed to dissipate. In Fig. 564
this is done by allowing the hot air to pass from the receiver D
through the injector nozzle F, and thus an additional quantity of
air is drawn into the cold receiver E to fill up the loss caused by
shrinkage during cooling. The air being compressed to 100 Ibs.
at a temperature of 484°, is reduced to 50 Ibs. in E, with a tem-
perature of 201° y but the gain is certain, for the heat has been
made to do work.

Much mechanical improvement has been introduced in the
compressors, such as the use of lever-lifted valves instead of air-
moved flaps, avoiding wire-drawing. Clearance spaces have been
much reduced, and the mains increased so as to bring the air
velocity below 30 ft. per sec. Referring now to Fig. 561, the
cylinders A A are compound high and low pressure, and the air
enters first the suction valves F F of the cylinder B. Leaving by the
valves E E, it passes by b to the surface condenser D, and then to
the second cylinder c, which it enters by H H and leaves by G G.
Finally it passes by pipe M to the storage receiver. The valves
are lifted by levers P, moved by cams N on shafts d d.

The Paris Compressed Air Company delivers about 8000 H.P.
from two central stations, through thirty-five miles of piping, the
further motor being 4! miles away. Prof. Kennedy measured the
efficiencies in 1889, and found that for one I.H.P. in central
engine, the customer received -39 I.H.P. with cold air and '47
with air re-heated just before entering his motor. With two-
stage compression and other improvements, Frangois shewed
in 1891 that a total efficiency of 46 could be reached with
cold air, '65 with hot air, and *8 if the hot air was sprayed
with water; which results have since been approached. This
assumes efficiencies of compressor, main, and motor at "9, '96, and
•93 respectively.

History of the Dynamo.


(12.) Hydraulic Transmission.—Refer to Chapters VII.
and XI.

(13.) Electric Transmission can only be briefly described
Faraday, in 1831, discovered magnetic induction, by which a
current is generated in a closed circuit wound on a bobbin, when
the latter is moved before the poles of a permanent magnet
(A, Fig. 565.) Pixii, Clarke, and others thereupon, in 1832,
devised the magneto electric battery B, where the bobbins are
rotated, and introduced the commutator to reverse the alternate
currents formed at c and thus ' straighten out' the total current.
Nollet, Van Malderen, and De Meritens improved this machine
up to the year 1871, dispensing with commutator, and thus pro-
ducing alternating currents (D). Dr. Siemens devised the H
armature E in 1857, working with compound magnet, and in
1866 Wilde employed a small Siemens machine F with commuta-
tor to excite the electro-magnets G of a much larger machine, and
thus avoid the necessity for large permanent magnets- The pro-
gress now was very rapid, and in 1867 Siemens, Wheatstone, and
Varley separately discovered the 'dynamo-electric principle/
by which the machine was made wholly self-exciting, the mere
residual magnetism in the soft iron core, whether new or after
use, being sufficient to commence the current, which then
gradually increased up to its maximum. K is a Siemens dynamo
with H armature and commutator, the currents being thereby

^..r^j* ^


4 ' •**  J»f '       *   / I   ,0 J, f, '  '» *• /   ix I

:<-   ,     •      r;      /<       i ... ;t    r.-

!   •  *  ',       /   /,«*,   f          ,     -U"

'    ' r  y t    •<       t f<        t     t <     >f j

7',                                          r    *       'f        / i    -   *            ,       >> i

4f * *-j  I

Jt   JV

-    *jf ">*>"/"


Hi      y

f r r
If  I ]



Direct or from Storage.                        551

of two sheets of lead a and ^ rolled into a spiral, with insulating
strips between, and placed in a vessel containing diliite sulphuric
acid. Charging till the positive surfaces were coated with lead
dioxide and the negative with metallic lead, the plates were in
such a chemical condition as to constitute a return battery.
Faure shortened the time of charging by coating the plates with
red lead (the lower oxide), and covering this with parchment tied
with strips. The only difference in action was that spongy lead

was formed at the negative plate, thus giving a large surface.
Present storage or secondary batteries (otherwise accumulators)
are on Plante* or Faure's principle, and do not really store
electricity, but change electrical energy into that of chemical
separation. They are useful where the demand for power is
intermittent^ and are fairly effective, the leakage, during a few
days being but smaU. (Seepp. 958 and 1118.)

Efficiency.—The work lost during transformation in a
dynamo may be as low as 8 %, though it more often reaches 15%
or 20 %. A*|redter loss usiialiy occurs, however, between generator

* ,4


5 ss


$ ; i


Electric Formula.

and motor, the resistance of the circuit causing dissipation of
energy as heat.    If

C = current in amperes,                Q = quantity in Coulombs,

E — electromotive force in volts,    W = work in foot pds.,
R = resistance in Ohms,


H.P =




746 R

v   -      R    -'
		v       R              E 746 H. P.          /Hp-.-76-p        w
o        w
		C          -VH.P.746R     .73?Q 746 H. P.              E2
		C2              746 H. P.

If 7 be the length of a circuit in feet, both lead and return,
and a — sectional area of wire in sq. in.,

R at 60°


a      1,000,000

when copper wire is used. Also if the E. M. F. drops from E to
e and the current from C to c in flowing from generator to
motor, and if W is the work put in by generator and w that
received by motor,

Efficiency of circuit =-- = - = —-

Example 54.—A dynamo driven by turbine can generate 50
amperes at 300 volts. The current is carried by two Np. 6 W* G.
copper wires to drive a workshop motor \ mile away. Assuming the
commercial efficiency of the generator as 86%, and that of the motor
as 84°/0, find the mechanical efficiency of the whole system*

Numerical Examples.


H. P. given out by generator = —7   =

H. P. Turbine must give to dynamo =

R of circuit, taking lead and return
(dia. of wire = "192, area = "03,
7=2640 ft.)

« -n   i         -         -

H.P. lost in wire =


300 x 50




2640 x 8*4
•03 x 1,000,000

jjo_xjo x 74

746 ~

= 20

= 23-25

74 ohms


H. P. delivered to motor

is   that   generated   less

=      20-2*48         = I7'52

that lost in wire.
H. P. available at shop shafting =   17*52 x '84
But H. P. given by turbine was 23*25


„          ~ .            H. P. taken out     14*71

Gross efficiency = —•--•--—-   -. — = ——

---------------------        H.P.put in        23*25

= '6327    or 63! %

and efficiency of circuit only •

H.P. delivered to motor
H. P. generated

= ££2? = -876   oi*87*6°/0
there being 12*4% of the generated H. P. lost in the wire.

Two actual cases may be quoted (i) 4^ H. P. was trans-
mitted 8 miles through •£%" telegraph wire, with a total efficiency
of 30 %. (2) The dynamos having a resistance of 470 ohms, and
the circuit 950 ohms, the line being 34 miles long, a total effi-
ciency of 32 % was obtained by decreasing revolutions from 2100
at generator to 1400 at motor, the potentials dropping 2400 to
1600 volts, a method of working first advised by Siemens.

(Seep. 929.)

Storage cells are objectionable for tramcar and locomotive
driving on account of their great weight, 2 tons of cells being
about the weight required for a i-ton car. The following results
are from an actual experiment with Faure accumulators :

35 cells of 95 Ibs. each
H. P. absorbed in charging
Time of charging
Lost work in charging
Chemically stored energy
Recovered electric energy


2 2 hrs. 45 mins.



60 % of 66 % = ,

!   £f*ft         rlf'j   -

**             /V   >

,t^'J^     <H$,v>/u.      j^|,   V*^*!1

^tW'J         ^^jr^'5-           4^.,'


(f«4/|T*l*|| *«ff   %^f
4*                *     J|f    '|J?|t$#

^»|*r«   ffiv4^^|^</|    ^*||   |^|

v*                  '*<  w»H*f   ^4|

>^v *? ^|   ^*f*

Solid Friction.


Laws of Friction.—* Solid' friction (or the friction be-
tween solid surfaces) is here meant, in contradistinction to fluid
friction. There are three laws, as follows :

The tractive force required to overcome friction :—

(1)  Depends directly on the pressure between the surfaces in contact.

(2)  Is independent of the extent of the pair of surfaces in contact, but (20)

increases in proportion to the number of pairs of surfaces.

(3)  Is independent (at low velocities) of the relative velocity of the surfaces.

Further, the force depends on the co-efficient of friction (JJL)
for the particular materials, thus,

Tractive force Fn = p. P    where P = total pressure.

	Method of Lubrication.
	Olive oil.
	Dry soap.
	Polished and greasy.

* Wood on Wood. . .
 Metal on Metal... Wood on Metal... Hemp on Wood... ' Leather on Iron. . Leather on Wood Stone on Stone... Stone on W.I. ... Wood on Stone...
	"5 •18 •6 '63 '•54 '47 7* •45 •6
 •65 •By
	•12 "I
	•21 *I '12
	•19 •II
	'35 "IS *I






As yu is the trigonometrical tangent of the friction angle </>
(seep. 560), the latter may be found as in Fig. 570, by dividing
a base-line into tenths and setting up p. on a perpendicular
from the mark i, to the same scale. Thus, for dry metal,
^ =* 10°, when /* = *i8. (See Appendix //., p. 868.)


Laws and Exceptions.

Morin's experiments are not true for very heavy loads or at
high velocities with much abrasion.    For the first, Ball gives

Fn = -9 + -266 P

for wood on wood, and the relation is set out in Fig. 571, the
dotted  line   shewing the result  of the  ordinary formula with

As regards velocity, at the Brighton brake trials, 1878, the
following results were obtained when the static coefficient was -242.

Vel. ft. per sec.
	ju between brake and wheel.
	fji between wheel and rail.






near rest

Solid and Fluid Friction.


As there was probably considerable abrasion in these results,
it is doubtful whether they should be accepted, further than
generally, for pure friction. The second column shews that the
wheels should not be allowed to skid when stopping the train.

If surfaces are thoroughly lubricated the frictional resistance
is of a ' mixed' kind, being neither solid nor fluid. The following
comparison is useful:


Fluid friction (gas or liquid) is:—

1.  Independent of pressure.

2.  Directly as wetted surface.

3.  Directly as v at creeping velocities.

as iP at moderate velocities,
as zr> at high velocities.

Solid friction is :—

1.  Directly as pressure.

2.  Independent of surface.

3.  Independent of velocity

(at low velocities).

The Friction of a Journal Bearing was investigated by
Beauchamp Tower for the Institute of Mechanical Engineers. The
load was carried on one braSs only, a top one, and the journal
ran in an oil bath. The coefficient varied with the lubricant.
With oil-bath lubrication Fn was independent of pressure, and

p, a -p.    In terms of velocity,         __


where c varies with the lubricant.   Thus, when v = 4 and/ = 300.

	C    '

Olive oil ......
	Sperm oil .


Lard oil   ......
	Rape oil...


Mineral grease
	Mineral oil

With syphon lubrication p = -^    where ^ = 2*02 for rape oil,

and with pad lubrication /* = *oi for rape oil.   The bearing seized
when/ rose above 600 Ibs.    (See Appendix //.,/. 870.)


Friction of Turning Pairs.

Friction of -a Collar Bearing. — This was examined
under the same auspices. Here the friction was nearer the
4 solid ' condition, the lubrication being less perfect. The
pressure/ varied from 15 to 90 Ibs., and v from 5 to 15 ft. per
sec. The coefficient was '036 for ordinary loads, the usual
formula being applicable. (See p. 871.)

Work Lost in Journal or Collar Friction. — R being
outside or mean radius respectively,

Work lost in foot pounds per m. = FnxV = ^Px2


H> P. lost = -

Work Lost in Pivot Friction. — Following the method

of Fig. 371, let r be the pivot radius in Fig. 572.    Assuming the

pressure to be equally distributed,

p                                               p

—- = pressure per sq. in,,  and £-5- = force of friction per sq. in.

irf^                                                         Trf^

Total friction on any ring = unit friction x area of ring

itP                      2LlP

.'. Total friction on outer ring == *--s x 2wr x ^ = "^ x f

and   Total friction on ring r% « — x -*— x f



the resistance increasing gradually from o to B c. But the force
must be multiplied by the arm to give the moment. The lamina
A B c D represents the moment for the outer ring, being

/2  IL P \

force ( -£— j  x arm (r)

Similarly ab cd\s> the moment at ring r2, and the pyramid volume
will give the total moment, thus :

2 u. P          r       2

Moment of friction =         x *• x - = - aPr

-----------------------        r           3      3r

If P and r are in Ibs. and inches, the moment is in pound inches,
and the distribution of pressure may be such as to reduce it to \
fjL Pr. Concentrating the total force at the outer ring, 'it will be

"^ and     Work lost per m.   =   f M P x 2 TrRN
which may decrease to £ /i P x 2 ?r R N

Example 55.—Find H.P. lost in a footstep, whose dia. is 4", total
load 3000 Ibs., revs. 100 per m., when /* = "06. (Hons. Mach. Constr.
Ex., 1887.)

H. P. lost =

2 X "06 X 3OOO X 2 X 22 X 2 X IOO

3 x 33000 x 7 x 12

Example 56. — Mean dia. of thrust bearing = 14", screw thrust
40,000 Ibs., and pitch 15 ft. /* = '003, and 1000 miles are travelled in
3^ days. Find H. P. lost in friction. (Eng. Ex., 1888.)

Speed of vessel

and as vessel travels 15 ft. per rev.


ft. per m.

Revs, per m. =

3*5 x 24x60 x 15

.-. HvP.lost=^--^^--«
-----------        33000


33000 x 7 x 12 x 3'5 x 24 x 60 x 15

i- The form of pivot surface may be flat, conical, spherical, or

specially formed.     If conical, -.— must be substituted for P,

where a = angle at cone apex.

(Seep. 507.)


Schiele's Pivot.

Vog. 573, is generated by a tracMx revolving
on its own axis.    The curve is drawn as follows : Step off equal

divisions i to 10; with radius OB and centre i set off ia and
join: with same radius and centre 2 set off i£ on 10 and join:
similarly 3*:, 4^, &c.; and then sketch the curve from B to K. This
pivot wears equally on all rings, but wastes more energy in friction:

Moment of friction
or 50% in excess of a flat pivot.

The Limiting Angle of Resistance^

a*perfectly smooth surface as at ., ^574, the


Angle of Friction.                           561

to the surface, but if the surface be rough, the reaction is inclined
to the normal by the friction angle, in a direction away from the
/#//P, and the latter must now be increased by Fn in order to
move the body. If not on the point of sliding, the obliquity of R
may be anything less, down to zero. Two cases are shewn for the
inclined plane, P being directed up or down the plane, but its
value may always be found by force diagram. In moving up the
plane, total pull must balance gravity + Fn, but in moving down
the plane must balance Fn - gravity.

Example 57.— A road engine weighs 12 tons. Find (i) tractive
force of engine to pull 48 tons behind it on a level road, and (2) the
load drawn up a I in 10 incline. Coefficient of traction =150 Ibs.
per ton.

(1)        Tractive force =(12 + 48) 150 = 9000 Ibs.

(2)        P x length = W x height     and P = W x -^
Also R will be found to = -995 W

Tractive force to balance gravity =       —~    = 224     Ibs. per ton.
Tractive force to overcome friction = 150 x '995 = 149*25 ,,         „
Total tractive force =                     373*25 „        „

But the engine only exerts 9000 Ibs.

.*. Total load on incline including engine = -— — = 24*11 tons.
and Load drawn exclusive of engine = 24*11 — 12 = 12*11 tons.

Diminishing Friction by Lubrication.—Spongy metals
like cast iron, brass, and white metal decrease frictional resistance
considerably, but the best results are only obtained by the appli-
cation of unguents.

Lubricants may be solid, as blacklead; semi-solid, as greases
and fats; and liquid, as oils. f Body' for support, and fluidity
to avoid resistance, are both essential requisites, and a careful
choice must be made between extremes. The following are the
unguents used for various purposes:—



%           4*1             tt                     #jt||               t|  4I {

* ^ I                         ftiw              M|t             jfft

!>#***!!    f'l'^l*^     '  |iJ#|A^|«/f                            ¥«^||r|Jf%f

|            4*| *«j« i   ju*f ty# jr*i|* Jf'iif/j*! t>^# i*| ?


r ,'    * * <t          *    ! I i   :          'is    J t i    ' i

-^        f    ^                            < ,»,     >      t (»*   j,|    v »• «•>

•   f   3     '•*    €      '*    «      -f '           i          i          It           *          . *

I             «      " * *           *              '     *     I         *   "    *          'I        » '           * *

t,             *             ,        ,    "        f   ,*  ,        8f  ,    ,     ,

• t       ,            *,r    f         «?r?r*   * •   *     <

^;»  tt   (,-*   I, ^ I f *

H       ;   •       '   1 >       4    i^J'lf

4 j..          / 4  >,   t   ,          * * w                         < t' > f* '  .   ;,   *   ' ?      '     * >    '< «* i   "     '

i ft          '.   I               ," i J i» f        V          '      •, ^ 4   - '« « J *   v    r,

,,      ;   a    f          .        ,      ^M                                         •'-•'     "**   t     ''.'^i

,; ^!    t ,*    |                                 '.**  ,|"I>^'U4 *

t *<*   !:*t
I I      *




I   I

The most effective test is obtained by machine, of which
Professor Thurston's (Fig. 578) is probably the best. A is a
pendulum hanging on the test journal B, whose brasses can be
adjusted for any pressure by turning the screw D E against the
spring c, while P shews the value, both totally and per square
inch. The thermometer G indicates the temperature. The
journal being rotated towards the right, the pendulum moves to
the left, together with pointer F, and the scale K at once indicates
the friction per sq. in. of journal, so that

__ P's graduation.

^ ~~ P's graduation.

Every five minutes during a test the revolutions, temperature, and
graduations are noted, values of p, afterwards found, and the results
plotted as curves wherever possible. In his ' railroad' machine,
Prof. Thurston used a full-sized locomotive journal.

Lubrication.—The oil-bath gives the best result, but is
rarely found in practice. The self-lubricating bearing, Fig. 5 79, is
perhaps the next best, where the oil is lifted by the shaft collar
and distributed by a wiper. The next in order is the oil pad, as
contained in the locomotive axle-box, Fig. 580, the bush merely
embracing the top half of the journal. Usually lubricators have
to be fitted, and are then designed for the conditions. B, Fig. 270,
p. 266, is a common syphon lubricator. The oil level being below
the syphon-pipe, a piece of cotton wick is placed in the latter and
hangs over in the oil. The fluid then rises by capillarity; and the
wick is to be removed when the machinery is stopped, otherwise
there is unnecessary loss of oil. Leuvain's needle lubricator, A,
Fig. 581, is a glass vessel, filled with oil, closed by a wooden plug
and inverted. Within the stopper a 'needle' fits freely, and the
oil trickles down the latter only when vibrated by the shaft. If
the dropping of the oil is to be observed and its regulation
obtained, such a lubricator as the Crosby sight-feed at B, Fig. 581,
may be adopted. When handle a is vertical, the valve <£ is
raised, and adjustment given by the nut d; but when a is horizontal,
b is closed, and the supply stopped

A loose pulley may be fed with tallow by means of Stauffer's
screw-down lubricator c. Oil would only fly away by centrifugal


3 ?*?:       f * i

* ?, * i "*± ? i -1"

*   41

^  ^      f

W* *     JJ

%      *           S

*     -%          1



1         f

Live Rollers, &c.
The equation becomes:

Q (8 4- gj) + W ^ == ? x 2 r


,-. p = Q(^ + SJ +


) is found by experiment, and

6 = "036" for rollers of wood 3' to 4' long.

b = '072" for rollers of wood i' long.

% = "016" to -oiS" for rollers of iron 5" or 6" long.


ft1;* 1»»>« t**t w

llftlMl"     -

^ifff-* ff1**1

V«*4        %f

If**   *«**

fiai*^l^!fv| U|             *» *jb

'"* f                     !     #?  t*     *»^**/^*

|«tv«|                 #  8|«ir*C^                  <4   i|

h*, f ws f^^UI *»r                iff                   t*»

lfn            4^^   #                         ^f 'if fe

;' %M                /^*/

tf                                                        «% *

*i*                     f^      ^f      *'4

il«*                                   fc* if«p                        ^ fj^f

»n*l                                                                         M            */f

; it - f   ;     ; ? r -  * %   *
5-l & 1L^ t      * ^ •? :  *       fc ** s  r   * v^  I

% t      11 r * r - - ^ > * ^ ^«* ••*

fY % « ^ T
» >JB^ *» i^


x   3

C   \

*   v


*ir   *»         *"    ^                   *

S         *

1W»        •*        ^

^     V    •*•


*   \  - t

»   I*    ^

? 8r i i


•.                >>      M

i *


-^ "*>

*  *



<  *  *T i

* * ? ;

*    *», *"

; c

gT   ^

i f

JS»   <*     t

c \


4   .^'v

|       *    III ti ^ |»rff

M I             J                            ^ -U--I-T ,-•


»|    f|


.:***•>?. X1

•   v1

1    (9

?      /**


Distribution of-Machine Friction.

Next plot these figures as in Fig. 595, the horizontal
shewing W, and the vertical ordinates the corresponding va
P; o D being the theoretical, and A B the practical profiles, "**"'

are both straight lines.    Draw AC || to o D.    Then at any ore!i
except o A the total P consists of:

1.  Force to overcome load, neglecting friction.

2.  Force to overcome friction of unloaded machine.

3.  Force to overcome friction due to load.

(2) being a constant quantity as shewn between lines AC, o f J

Then, if iv be the equivalent weight of the unloaded
causing friction,

(i)          (2)        (3)

F            BC

From (3) we find u = -~ =----------= '00687.

r     W      5 x 2240             '

Supposing p a constant throughout the machine,


p «----+ -00687 w 4- -00687 W


and as CA = 23*1 = 3362 x -00687     :    w — 3362 Ibs.






2-, i * • I r : i I *

%» > ^ **

% # r r **

! f.

> *

*»      m

*    *

"      *


Transmission Dynamometer.

Friction is only the medium for absorption, and does not enter
into the calculation. The arrangement at K permits adjustment
for various motors.

White's Transmission Dynamometer is represented
in Fig. 597. A is the, motor shaft, and B that of the driven
machine. As A turns left-handed, the arm F E is held back by the
weight E, and thus B is turned to the right. Supposing the arm
were carried round, no work would be given to B, which would be
stationary, but E'S rotations would be half those of A (see fig. 526),
The load supported on A would therefore be half that on E (at
equal radii). But although the power be taken off at B, A and E
have yet the same relation, so that

List of Efficiencies.                           577

work transmitted = load on A x distance travelled on A
and load on A = half that on E,

1                 •            E               T)   XT                 A   IT    t,         E 7T R N

.-. work per mm. == — x 27rRN        and H. P. =

F counterbalances the lever weight.

List   of  Efficiencies. — Efficiencies   in   various   cases, as

found by the methods previously described, are  as follows :

Cranes worked by spur gearing...        ...        ...        ...    30% to 6o°/0

Worm gearing (indifferently constructed)       ...        ...    30%

Worm gearing (very carefully constructed)     ...        ...    9O°/0

Weston pulley block, well greased      .........    30% to 4O°/0

Screw jack          ..................    iS°/o to 35%

Cornish engine(Brake H. P. -f Indie. H.P.) ......    35°/0 to 6o°/0

Other engines (Marine, Loco, Gas, Oil, &c.) ......    75°/0 to ^5°/0

Undershot water wheels ...............    25% to 30%

Overshot water wheels  ...... •      .........    7°7o to 75°/o

Breast wheels (Poncelet floats)            .........    6o°/0 to 65%

Pelton water wheel        ...............    8o°/B to 9O*/0

Turbines (foil .sluice)      ...............    6o°/0 to 8o°/0

Hydraulic press (neglecting pump)      ...        ...        ...    9^°/o t° 99%

Hydraulic jack with pump       ......                  ...    77%

Pumps (piston)    ..         ......        .........    7&%

Hydraulic accumulator ...        ...        ...        ...        • • •    9 * %

Hydraulic lift, working rapidly           .........    50%

Hydraulic cranes, all losses taken       .........    55%

Mechanical efficiency of engines, not varying appreciably with
load, is often found by comparing an indicator diagram taken
' running light/ with that under working conditions. (See Ap^
pendix IV., p. 962.) (See pp. 874 and 1125.)

We will close this chapter with a few comparisons of power


Advantages.                                     Disadvantages.


Useful in modifying power and
obtaining special motions, as with
valve gear, parallel motions, &c.
Coupling rods a case of pure tians-


Frictional loss slight

Dead points often occur^ to be over-
come by force or chain closure.

Will only transmit over very short

but depends on number of joints.


4   'f «»M« f*r

f<ff*t^   (,*#**§*   «S    *'^                                        V                        f     »f

MJ    •**   /I?;    "**"*    *f/   ,****    'fri»0
'«M*     ,/^iA    ^n,*»*  * 4* *****   %$ff Jit    ^*r//*l##       **t                  fl|t      f*

ffwl      ^*«|<«               f j>*».< f    f»'*fa|

^ll'l        * ^!ir«   "'^ * ,/B^ f**'J$  j,'**!»*ffr^irt|(|            W*+   ^ifttf    fti^?f#«              t   t*»?*|

4«|    >-K^

"                  1*1^1*1

#P                   4|

** 's


Tension Elements.





Useful in connection with shafting
as a distributor and modifier with
comparatively few parts.

Easily started and stopped.

Practically noiseless.

Very convenient for bridging rea-
sonable distances.

Large pull on bearings, but in well-
lubricated bearings friction does not
depend on pressure.

Slip an advantage in case of shock.

Frictional loss principally in the
line shafting: about 257,, to 50% in
a shop system.

Large belts with heavy pressures
are expensive to maintain.

Slip   a  disadvantage   where  exact
velocity ratio is required.


For fairly long-distance driving in
mills, and for travelling cranes.

Better grip than belts, due to wedge-
groove pulleys.

Quite noiseless.

Separate driving to the various
floors of a mill occasions less loss of
time in breakdowns.

Small liability to break down also.

Frictional and other losses probably
somewhat larger than with belt gear-
ing, due to heavy pullies and fly-

Working speeds being high, rope
tension is increased 5o°/0 by centri-
fugal force : but bearing pressures are
not thereby affected.


Suitable for very long distances, say
for several miles, when relays are
adopted. Cases quoted in text.

If moderate speeds be employed,
little increased tension from centri-
fugal force.

Frictional and other losses 22S/0
per mile, not including motor and
machines: lesser and greater distances
in proportion.


As useful as belt driving in de-
creasing the number of parts while
modifying the power: but gives at
the same time positive transmission,
and may be used with heavy loads.

Adapted for high as well as low
speeds if well made, but the former
should go with light pressures.

Frictional loss depends very much
on design and manufacture, and pro-
bably varies from 5% to 3O°/0 in a
pair of wheels: there being two sets
of friction surfaces, not including the

Increase of pitch after wear causes
excessive friction and bad working.


Other Transmitters.



Disadiian tages.

Almost noiseless and n on -vibrating.
Advantage of slip when shocks are
Useful for high speeds.

Frictional loss about the same as
for belt driving to shafting, but
comparatively small with one pair of
wheels. Unequal wear.

Large pressure on bearings ; de-
creased in nest gearing.


Of great value for long-distance
transmission in close workings.

Better than hydraulics when high
speeds are required in piston motors.

Loss by cooling varies from 70%
under bad conditions to 20% with
re-heating and air injection.

Loss per mile by friction about 5°/0.


Suitable for Jong distances.    More  j
especially useful for intermittent de-
mand in power distribution, and the
concentration of immense power by
aggregating storage.

Leakage slight.

Inertia an advantage sometimes, as
in riveting machines.

Losses slight if low velocities are
taken, say I5°/0 in usual machines;
5°/o per mile due to friction in pipe.


Unsuitable for continuous work.

Uneconomical with high velocities
and reversible motion, on account of
shock due to inertia. (Damage ob-
viated by relief valves.)

Velocity should be kept down to 4
or 6 ft. per second usually, and slow
moving rams adopted, necessitating
multiplying gearing.

Piston engines run at 60 or 80 ft.
per m. but are usually wasteful.


Especially suitable for long dis-

Wires may be conducted in any

No moving parts in line of trans-

Easy subdivision of power.

May be stored by secondary cells.

Loss in line varies as the square of
the current used (C2 R): hence high
voltage is adopted for long lines,
giving an economic loss of from 5%
to 40% in the line.

Loss in dynamos from 5% to 20%
each, of the energy intrusted to them.

Storage cells, being heavy, a*e
not really suitable for transportation
purposes. Loss in charging and
discharging, say 50%

(See Appendix //., p. 875, and Appendix III,, p. 928.)

X    t



Radiation and Conduction.

both have equal temperatures.    Such transference may occur by
radiation, conduction, or convection.

Radiation is the passage of heat between substances not in
contact, without at the same time raising the temperature of the
intervening medium. Thus a fire may heat surrounding solids,
and the air receive its heat from the solids in turn. To explain
radiant heat, a fluid of infinite tenuity is imagined, called the
Ether, filling space and the interstices of matter, and transmitting
radiant heat, by wave motion, without increasing molecular
motion. If, however, the undulations be arrested, the energy
is absorbed as molecular motion, and becomes apparent in the
arresting body as heat. Radiation is an aid to heat dispersion,
as in heating apparatus, but a disadvantage with boilers and
steam cylinders, there causing loss. Good radiators must
therefore be adopted in the former, and bad ones for the
latter case. Good radiators are good absorbers, to an equal
degree, and reflecting power is the exact inverse of radiating



Lampblack or soot
Cast iron, polished
Wrought iron, polished
Steel, polished ...
Brass, polished ...
Copper, polished...
Silver, polished ...

Radiating Value.





*    5

Conduction is the transfer of heat by contact, molecular
motion being then directly caused. Heat is thus transmitted
through the thickness of a furnace tube. There are good and
bad conductors, the former being chosen for fireboxes, other
properties being suitable. (See Appendix /Z, /. 876,)



	Relative Conducting Value.



	2 7 "I

	*O  A II'O

Steel        ....... ..
	1 1 '6


Bismuth  ...

Water      ............


Bad   conductors   are   of value   for   clothing   boilers,   steam
cylinders and pipes, &c.    (See p. 902.)


	Relative Obstructing Value.

Silicate cotton or slag wool Hair felt ............ Cotton wool Sheep's wool      ......... Infusorial earth  ......... Charcoal ...
 35H 82
 73'5 73*5
 71" &

Sawdust   ...        ...        .......

Gasworks breeze ......... Wood, and air space
	43*4 357

Convection is a means of transmitting heat to liquids and
gases. A flask of water being placed over some heat source,
the lower or heated portion of the water becomes lighter and
rises to the surface, up the vertical centre-line, only to become
cool again and flow down the sides to the bottom. Thus are
continuous (convection' currents formed, which soon distribute
heat throughout the liquid. Similarly also is the air of a room


Expansion and Temperatiire.

heated : the fire, near the floor, rarefies the immediately sur-
rounding air, which rises to the ceiling and falls again when
cooled against the walls. Water, being a bad conductor, cannot
well be heated by any but the convection method, hence the
adoption of ~a low position, in a boiler, for the fire-grate.

Expansion is the result of the application of heat to all
bodies, whether solid, liquid, or gaseous; the first being least
and the last most expansible. Many examples may be suggested
of the application of this law, some useful and some detrimental.
Shrinking of gun coils is of the former type, while the endlong
clearance between rail lengths of the permanent way avoids the
injurious effects of the summer heat. Fig. 327 shews how work
might be done by the expansion of solids. Water, between
32° and 39'i° F., is an exception to the law of expansion; during
that period it contracts as the temperature increases. Cast iron
also expands when solidifying in the mould, and bismuth and
antimony follow the same rule; gold, silver, and copper contract.

Measurement of Heat.—We proceed to measure intensity
and quantity of heat, bearing in mind, however, that heat is not a
substance but a form of energy.                                          !

Temperature is a measure of the intensity of heat, being
registered on a thermometer or pyrometer. Thermometer^ are
based on the expansion of liquids or gases in a glass bulb, which
then rise in a capillary stem from which air has been exhausted.
Mercury or alcohol are the usual liquids, the former for ordinary
and comparatively high temperatures, and the latter for very low
temperatures : the boiling point of mercury being very high, and
the freezing point of alcohol very low. The freezing and boiling
points of water, under atmospheric pressure, being unchangeable,
are first marked on all thermometers, after which the graduations
are spaced according to one t>f the following methods :

	Divisions between Freezing and Boiling.
	Freezing Point
 Boiling Point,  j

Fahrenheit ... Centigrade ... Reaumur   ...
	1 80
	21.2° 100*


Quantity of Heat.                              585

Reaumur divisions are adopted in Russia; those of the Centi-
grade by scientists and the Continental public \ while Fahrenheit
divisions, being used by English engineers and the English
speaking public generally, will therefore be adopted in this work,
and the Fahrenheit degree be looked upon as the unit of intensity.
Centigrade readings can be translated into Fahrenheit and vice
versfr, by the following simple formulae:

F = (C° x |) + 32     and C° = (F - 32)$,

Pyrometers are required to measure excessive temperatures,
such as those of furnaces; they are discussed on page 587.

Air thermometers are of advantage in experiments of great
delicacy, because small increase of heat will cause large expansion
of air. The instrument is usually laid horizontally, and has a
small index of coloured sulphuric acid, as at c, Fig. 601, which is
moved along the tube by the expanding air, the end B being
open to the atmosphere. The reading is considerably affected
by change of atmospheric pressure, so the barometer reading
must always be taken, and a correction made to standard
pressure. The expansion of gases is more perfect than that of
liquids. (Seepp. 587 and 1126.)

Quantity of Heat.—More or less heat motion may exist
in a body, depending on mass, heat capacity, and temperature.
The British Thermal Unit (B.T>U.) is the amount of heat required
to raise the temperature of a pound of water through one Fahrenheit
degree, the water being near its greatest density 39*1° F. This unit
represents an amount of energy equal to about 772 foot pounds.

Specific Heat.—But some bodies have greater capacity for
heat than others, that is, weight for weight, will absorb more heat
for a definite rise of temperature. Taking capacity for water as
i9 the relative capacity of another substance called its Specific
Heat^ is therefore the amount of heat in thermal units required
to raise the temperature of a pound of the substance through \)ne
degree F. Bunsen's ice calorimeter has been used to determine
various specific heats, but we shall describe the method of mixture,
which is precisely the same in principle. The body, being regu-
larly heated in a bath of steam, is removed, and put in a vessel
containing a measured weight of water at .& certain,


Method of Mixture.

When the body and the water are in thermal  equilibrium, the

final temperature T° of the mixture is taken.    Then, the heat

lost by the body divides itself between the  water and the
casing, so

Heat lost by body           = Heat gained by water & casing,

weight x sp. ht. x fall of temp.   = weights x sp. hts. x rise of temp.
ws (/• - T°)               -        (wl s} -f wfy) (T - /!>,

where w wl and w^ are the weights in Ibs., and s sl and J2 the
specific heats of the body, the water, and the casing respectively :
jj being unity. The value w% s% is known as the * water equivalent'
of the vessel. The first temperature of the water is // while f is
that of the body after steaming; and the changes are shewn
graphically at A, Fig. 598. Inserting known values, that of

s may be found, the following table being obtained by this and
other methods:—


Water at 39 'i°
	...   I '00
	Wrought iron
	...    ^113

Water at 212°
	... 1-013
	Steel ......
	...    -116

Ice at       32°
	...    -504
	...    -095

Steam at 212*
	...    -48
	Coal ......
	...    '24

	...    '033
	Air    ...    ...
	...    -238

Cast iron
	...    -13
	Hydrogen ...
	... 3"4°4

Water Pyrometer.


Example 58. — Find the specific heat of copper from the following
data : — Half a pound of copper is heated to 212°, and being plunged
into a pint (20 02.) of water at 60° contained in a wrought iron vessel
weighing- 4 oz., raises the temperature of the latter to 65 f°,

(«/,*! +. w±s2) (T - O _ (1*25 + "25 x -113) (65! - 60)
J=       """^(/°-~T0)           -         ..... •5(2i2~-~65|) ......          =  °938

Pyrometers, for measuring very high temperatures. — Wedge-
wood's and DanielPs, based on expansion of solids, are now
obsolete. Siemens' electric pyrometer measures the resistance
of a circuit, which varies directly as the temperature of the wire.
Wilson's and Siemens' water pyrometers depend on the method
of mixtures. A cylindrical vessel of sheet copper, clothed with
felt to retain heat, is provided with a cover and thermometer (see
B, Fig. 598). A small solid cylinder of copper, of known weight,
being placed in the furnace whose temperature is required, is
shortly removed, and plunged into the water of the pyrometer,
^vhen the latter is closed. The final temperature of the pyrometer
-water being observed, that of the furnace can be deduced. (See
Appendix //.,/, 876.)

Example 59. — Find a flue temperature by water pyrometer
from the following data: — Quantity of water = I pint, its first tem-
perature 65° ; weight of copper cylinder — 4 J oz, ; final temperature of

•water = 72° : water equivalent of vessel = '38 (Ib. F°)

+.:3.8) .

4- wsl0


__                                                   2 5 ± '38) 7 H- C'26s x -095 x 72)

ws                                        "265 x '095

Expansion of Gases.—Two laws govern the varying
volume of a gas, according to whether temperature or pressure
be kept constant,

The first law of gas expansion, discovered by Boyle ia 1662,
and verified by Marriotte in 1676, states that the volume of a given
portion vj gas varies inversely as its pressuref if the temperature be
constant. Shewa by symbols:

V oc =     and PV » a constant.

The relation of P and V is given by diagram in Fig, 599,
the ordinates PPX of the curve representing pressures, and the

1                        '                          RR

Boyle's Law.

abscissae V VI corresponding volumes, a temperature t being
maintained. Only one curve, the rectangular hyperbola, has
ordiriate x abscissa constant throughout, and that is the form
of the curve AB, Although always approaching the co-ordinates
oc, OD, it only meets them at infinity.

Isothermals.—By reason of equality of temperature, AB is
also known as the isothermal of a perfect gas, that Is, of a gas
following Boyle's law perfectly. Marriotte's tubes, Pig. 600,
prove fairly well the accuracy of this law* A Is a closed aa<i a



*   ^

*    f-

i*   5 !

~»t                         ^         "

%     t t

-                v ii   T

^     * *      r *  »


*= .    : s 1        t

- * »    * f t   ,*  J

£,;  t "" * f *   -    -1

*~*2».1r*»*                ?

^ -1f I s  ; I

I        £ > f ; |

3^       il|l| £   2


* *       * 4i£ I

&   {;    -     ^   s

J   %

Latent Heat of Water.


The above formula gives P or V*at any temperature, when c
is known.

Three States of Matter.—These, the solid, liquid, and
gaseous, are well understood, and it is also now admitted that all
bodies are capable of existence in each state successively, though
not necessarily at the ordinary pressure and temperature. Taking
one pound of any substance and applying the specific heat due to
its state, its temperature rises one degree, and as the specific heat
is approximately regular for each state, practically the whole heat
is registered on the thermometer. But in all substances two
critical points occur called the points of fusion and evaporation,
and known respectively in the case of water as the (freezing and
boiling points;7 at these points additional heat is absorbed merely
to do the work of re-arranging the molecules, of fusing or melting
on the one hand, and of evaporating on the other hand. Such
'latent' heat is not observable on the thermometer, and must,
therefore, be otherwise detected.

Latent Heat is the quantity of heat units absorbed or given
out in changing one found of a substance from one state to another
without altering its temperature. This phenomenon, first observed
by Black about 1757, will now be demonstrated in the case of
water, and the units measured.

Latent Heat of Water is that required to melt one pound
of ice at 32° F. Provide a vessel with felt-covered sides, similar to
that at Fig. 598. Fill it with water of known weight (w) and tempera-
ture (t°). Take a piece of ice which has begun to melt, wipe dry,
weigh (0/j), place in the water, and close the apparatus. When
the ice is quite melted^ gently stir, and measure the final tempera-
ture (T°), which may be a few degrees above 32°. Let Lh = the
latent heat of water; then

Heat lost by water   =             Heat gained by ice

weight x fall of temp. « weight x (latent ht. + rise of temp.)
w(f-T)         -          ^{Lh+ CT-sO}-

Supposing 20 oz. of water at commencement, at 60*, and 2 oz,
of ice at, of course, 32°; the final temperature being 45°, then

20 (60 - 45) - 2 (Lh + 45 - 32)»

, r        300 - 26                .

and LI* ** a—__ « 137 units. •

592                      ^Latent Heat of Steam.

Supposing, further, that one degree in the fin.it 1    tr

has been gained by radiation from the  room,    -4-4*     **
temperature due to mixture,

20 (60 - 44) = 2 (Lh + 44 - 32)

The correct result should be 144 units, only    ol>f *%
careful preliminary radiation experiments.

Latent Heat of Steam.—In Fig. 602 wat<sr i ,
flask A, and steam then passed by tube B to flask c:, \% I.
into water. The screen D is to prevent radiation- fir**?

and the experiment is continued till the water ne£irl>    f "
Weighing c both  before  (w-^ and  after the
difference is the amount of steam condensed (w).

Heat lost by steam = Heat gained by
•w {(212 + Lh) - T\ - wl (T - ^).

Suppose the weight of water is 20 oz. at temjje?imt
weight of steam condensed i|o£., and the ^nal
a loss of i° occurring by radiation,


~ 147)


TJi$ exact value is 966 units,

. .    It should be well grasped 'that latent heat is a                «if

teat given to the ^A^^ti^^ solid' Co'

't't't'    <*    I f 1    [fv  *

'(  .»*;    f n |   v|

|«i,|if   'HfjflPf*

i,     ,     •  ' «^ j *'/

594               Saturated and Superheated Steam.

the volume of i Ib. weight is 26*36 cubic ft, termed   *
volume : and the latter always = relative volume x •016*

Def. i. — The Saturation Point is attained iv*'

latent heat required for the steam has been  f " '

Def. 2. — The Boiling Point occurs when the tetf   *
water overcomes the surrounding pressure.

Def. 3. — Dry Saturated Stearn is that which k<* *
volume, pressure, and temperature, correspMt* * ;
complete formation.

Def. 4, — Wet Saturated Steam is in process </. ^
and is in contact with the water.

Def. 5. — Superheated Steam is that which h** '
perature raised above formation point.

Def. 6. — Specific Volume is the number of cttbi*

Ib. weight, and SPECIFIC DENSITY is the nuws?"
in a cubic ft.    (See pp. 766^^933.)

Dryness Fraction. — If the weight of water pat*11
given volume of wet steam be measured by suitable   ^
the proportionate wetness will be shewn  when  that    **-*
divided by the total weight of dry steam and water   4
while the proportionate dryness, or

TA            r     .•               weight of dry steam         (See Appen* f

Dryness fraction - -————-                ^

f  :                        thermometer,

Curves   of   Saturation   Points. — The   cont| «,!**"*'
temperature and pressure of  dry saturated   steam   I  * *
proved by experiment    From - 22° to 32°, Gay-Lus»»f    **«/**
apparatus in Fig. 604*   Both barometer tubes have vat *  *
the mercury, but B has a, little water on the surface of the*
whose vapour pressure reduces the height of the
i in. of mercury represents about •£ Ib. per sq. in., the          -w

therefore known.  Various freezing mixtures successivelf* * «^v,i
the blind end of tube 3, their temperature being

Fig. 605 was Regnaulfs apparatus for temperature*
to 122*.   As before, barometer A bus* a perfect
B*S vacuum is impaired by vapour,, rising from water              4


/*/ /



JiC *!


li 12

^ •»  I          4* * 4 »   /*«


Regnaulfs Experiments.

surface of the mercury.    Both tubes are surrounded by heated
water, whose temperature is shown by thermometer.

Regnault further found, as in Fig.-606, the temperatures and
pressures of saturated steam between 122° and 219°, the experi-

ment having since been carried to 432°. A is a boiler where
steam is formed, and B a copper sphere containing an artificial
atmosphere, produced by the condensing syringe c. As fast as
steam is formed, it is condensed by water passing from D to E
round the steam pipe; but this is a practical detail Essentially
the pressures in B and A balance, being measured by the open

«*\                                   *.'   *'-'    '

^  ?

If f

I   If         4|'*   H.14vl*i   I

; t

I If

*** i



*       H      f


i i



«*   f * t*

-* f J^
S1    3 %







I A   * ** ;  *«'.                                                                                                                                                          *$£«,   '<$

f^it  • *   »"<$#'jj                   If   il-J^f^f  '.f^^f |        ^*                                     '«*'*   pit**'¥***$

tA^l* I   'jj*'*^*    ^'/    ft/a    %jt"fy***(tl    '*   ^    tf   t^ffll'ltli*

f           « ,11*' y   ,^|^     f IS I   "f  ,   j<^*t**   **M«»»'   '*  '*lfr*VlH  fk  v  ^         I*'  pti.^1%*^

^^|  .J*'!*'! cr|   |« J*^^./', <*4    ^^!   m    f'jfr   |f««#$^«f*rfffS    ^*'*4                 f*'||   i^     J«ifjr|:|

^f,^^'^*    r ,«'*    |i   ?%f^t*^   -f    w»    /l^i*^   /,*!><;,    /ifrl   I      * i'^i1           %trf|^|"t^

fjr/M                    /*    J   /*   „

"          *m





$ Jf * I/1
I              '



First Law of Thermodynamics.

From the circumferences of F F weights G G were hung, %% !
being allowed to fall, rotated the paddles and raised tb«-* !
perature of the water. By repetition, the temperature * ^ r
water was raised to a measurable quantity, the work of the ** H
weights being simultaneously noted, until the average of f Sl
experiments gave the ' mechanical equivalent' as 772 foot 11 *'**
one British Thermal Unit. We may now state the
First Law of Thermodynamics. — Heat
energy are mutually convertible, and Joule's equivalent (f)
rate of exchange. (See pp. 930 and 1130.)

Internal and External Work during Evapormfci

—In heating water and evaporating it:

i.—The temperature of the water has been raised.

2.—The water has been changed into steam at the    *;

3*—The volume of the water and steam has been

against external resistance.

* Rowland's later value, 778, is probably more nearly correct*

f I-'   *    ^ *    | . \t ^      «         .1

f      ^   f         A        i    -         !,    |         '   ^      , "       »J     '

i       \ '»' ))  ^ f i %r


Efficiency of Steam.

Commencing at o, Fig. 611, draw the co-ordinates ov, ox, for
pressure and distance respectively. Measure 26-36 ft. at OA, and
2116-8 Ibs. at OB; the rectangle AB then shews external work.
Make OD and D E 12" 36 and 2'i times o B respectively; the area
o F is the internal work during evaporation, and D G shews the
work required to raise the water's temperature from 60° to 212°.
Rectangle AB represents the only useful effect, the rest being
expended on internal changes, and the

external work

^~. .          ,. ,

Efficiency of the steam

- - -


------- r     .-     =    *'    ^

total work         863,096

= '0646

— -

Let us next examine the case of steam at 160 Ibs. pressure
(above atmosphere), as in triple-expansion engines.

i Ib. of steam at 174*7 Ibs. per sq. in. absolute has a specific
volume of     ..........      2*5        cub. ft.

Load on piston = 174*7 x 144    ..... = 25,156   Ibs.

(1)  External work = 25,156 x 2-5     .    .    .    . = 62,890   ft. Ibs.
Temperature of steam     ...,...«= 370°      F.
Latent heat = (966 - "7(370 - 212)} x 772 = 660,369 ft. Ibs.

(2)  Internal work = (660,369 - 62,890)     .    . = 597, 479 ft. Ibs.

(3)  Heat to raise water's temperature

= (370-60)772 = 239,320 ft. Ibs.
And total work = (i) -H (2) + (3)   .    .    . = 899,689 ft. Ibs.

Efficiency of steam

external work _^  62,890
total work    " 899,189


Which proves that high pressure steam, weight for weight and
without expansion^ is not more economical than low pressure

Specific Heats of a Gas.—As with solids and liquids
these are the quantity of heat required to raise the temperature
of i Ib. weight through one degree F. But there are two methods
of raising the temperature, the specific heat being a different
quantity for each case. Assuming the gas enclosed in a cylinder
and covered with a loose piston, we may, while supplying heat,
(i) allow the piston to rise freely, or (2) fix it immovably. In

™*~ are heating at constant pressure, and in (2) at constant

If trjfa p*                    **w»

»->',                          f              *    'f«J*«^4*![      ^4Pjt***                    ^f|                         '***

t^h^f    M"*M    *   i*   *'>% |                              /f*   fl*            *|J   #*

,***   |f   /r*   **   |^                     iff/

fj^'^tJ ,*'* j >fr "?*   *-*   *J*   *           ** I4f|

Ifr^  I.   if  *«}    ft   f^ti     -    *//;      *             w,1

f /* *.«if*             4

.*    I ,  I         I   f

f               § *    '


I *     *   "  5     <t   *' i «


Specific Heats of Steam.

Let us heat a gas under a constant pressure P,
being increased from Vl to V2 and the temperature   ri**1
T-J to r2 absolute : then

External work * P (V2 - Vj) = c (r2 - r- , >
Total heat expended = spec, ht x rise in tern } **

' = Kp (r2 - TJ)
and Internal work = Total - External

= Kp (r, - T-J) - * (r2 -    *-» I

But, when a gas is heated at constant 'volume,
work is done.

Kv (r2 —

KP (r, - rj -


and Kp - K

Note that internal work is always Kv (final
temp.), and may therefore be positive, negative, or

Specific Heats of Superheated Steam.—By
Kp = 370*56 foot Ibs., and as steam a few degrees al>**v*
tion point is a practically perfect gas, Kp will be a regal        4
Further, if we are heating at constant pressure,

For steam PVS = csr               For air PVa =» <r^ ur

Now the ratio of specific volumes —




Then Kp - Kv » 85*5 and Kv
Finally y = —£

•622     '622

- 37o:56 =

~ 285*06     •=-^:

Expansion Curves and their Areas. — The

co-ordinatdng -Boyle?s law, has been shewn at Fig,

other expansion curve, as these are called, has the forr» it H

the exponent n changing with the substance*    Now

area, Fig. 612, shews the work done during expansion^   mi

Isothermal* and Adiabatics.

be actually measured (see Fig. 325); but as these curves have
definite formulae, it is easier to use algebraic methods.    Then,


Area of curve having formula PV = C        is PV x loge ,—


and as PV = cr. and — = the ratio of expansion r.


Area = cr loge r.              (Seep. 1131.)
(Use hyperbolic logarithms, and see Fig. 612)
Area of curve having formula PVW = C       is ——------?—?

71 — I

Isothermals and Adiabatics.—If a gas expand, and
advance a piston against a resistance, it does work requiring
expenditure of heat. Such heat being abstracted from the gas,
the temperature of the latter falls; but if heat be supplied just
as fast as it is abstracted, viz. equal to the work done, the tem-
perature will remain constant, the expansion be according to
Boyle (PV =» C), and the curve be called an isothermal.

If no heat be supplied, the pressure-volume curve will fall
below the hyperbola, as in Fig. 613, according to the formula
PV" = C, and be then termed an adiatiatic. Similarly, in com-
pressing, the adiabatic will rise above the isothermal, because
the gas becomes t heated by work done upon it (Fig. 614).
(See Appendix IL,p. 879.)

Adiabatic Exponent.—The value of n will now be found
for the adiabatic.

Area of curve

n _


External work,

Total work = Internal work + External work

- <- - '->    •-•-•>_+,K-K') - h - o (SSf

* Notice change of sign in two places in order to balance.


***        C

»      -     *       ^*^*»»

«*    *c~       *    %r       *    *t      v    J|    ^

i        *J


V      **      *^

f   ^     i   J      I     f

m  *    -    *    i-   *.

I r s  «   * f   i 5

3t        t        ^ ^       **^


f f,

•/? "J *, *

/if ,'»


f^ij**/ I*, r «t»,   A' "*^     .*11   '    4»{ **    4!> fc* 4,*  *    "    *    "* *

tlf                                          *^l

Jfl 1 »|f    M',   * *   *i 1'<»r   4i»,i4f     *i        .'^    *   '/   /   ^   ^-    ^'   l/^» ,»>.|.'

*'&«*««>»    «1    |/ff ^J  '«*    f       «         IM r     *      S />   *j       ,   f!*?^.     //'' **"|

Ir^f»|/Cf^1ijff           tl    *     t^"     tft'irJKWit      'f4     **    ^JtiC'M «».' >      ** <f.   ,«<«^'r*r

f o|T,t?*^t,   t^«,'.f   tif*i»,   l^^.r^/M,   f  /i *i"j.|   '.^ we, 4   ^/^^  04

t^ I*,  t.'^t,f   ,«   4   '     l^r   il* A^      V * « rv, * 4   v*^/       I*   §£*)>*    *

|#            ***<'•*   *'5 *i'"!*i *  *    *'*"'*   *'is<*ie   4hf   t,«! ,.*it*'    *•      vftf

}f   *'t    i   <   <»jf*<    4  I     1    *«"/<.?  fc  1t   <"', **'**,»W'*«'          '/ *   *     '»   tt^   /•'kiflAfit

il*'*'   r |f i a ^4  >'*V   '"ivf**    J ^   /f *   ?*«'?  f*jfitf$  tit totn         *fe   **•/   ', ,flif|
fff %*fl|^M-f%   fi f   !lrr   Hrit^a   ,|   »*df   fjt|' #f   f/*,/    |«7*i*4 f       i't'*    M!/*

f f li^t sW                ^^ '* ** «    « ^V* f*^^ i      , ;»^M /   **#* »•

til                     II *           g*i I*

i           <4                                                           i^ ,4*

if^                             t

i#*# ft%                                                           tit                            ^^v

f                       *|,r^


it <Hr                                       «| sn              *t«v     t< tl '^ M u

Ilir             M|                             "«r|rf#*, f f*s   * i^

w          it4i»*4^ i              f^jif/* ft

¥«^   Jfi^ffll*,  ^fl  |i»   |*f*|-if^^v              ^tf

*l, i#A<*&t*   Iff  |<t#«<M#S  f^|^i-ri        ^,'j)         !»^§t i

t^tr^   4^                          «

w| ?|K                                                                     IM

Conditions of a Perfect Heat Engine.


mechanism; and, finally, reject a smaller quantity of heat into
some cold body. In the steam engine these ' bodies3 are the boiler
and condenser respectively. We shall see that the efficiency of
the engine does not depend on the working substance, if a rever-
sible cycle be adopted, but only on the difference of temperatures
between which the substance is utilised. A perfect heat engine
should have the following qualifications :—

1.   The heat must be received at the temperature of the

hot body.

2.   The healt must be rejected at the temperature of the

cold body.

3.   The cycle must be reversible.

For perfect working, it is clear that all heat represented by
drop of temperature between the hot and cold bodies should
be delivered to the engine as work. But if there be a fall of
temperature between hot body and engine, or between engine
and cold body, some heat will be lost on the way which does
not reach the engine. Hence the reason for (i) and (2). We
may explain (3) similarly, first premising that by direct action we
mean the transformation of heat into work by abstraction of heat
from hot body; reversed action being obtained by turning the
engine backward, giving all the work'back to .the hot body. In
a perfect engine, the work given by the gas during one direct
cycle must equal the heat returned during one reversed cycle,
which is to say, that all the (available' heat must be transformed
into work.

Carnot's cycle fulfils these three conditions, and none other
can have a higher efficiency, as we shall prove. Fig. 617 is the
ideal engine, having a non-conducting cylinder A, and piston B,
the latter connected to suitable working mechanism, c is the
hot body, E the cold body, and D a non-conducting cylinder-
cover; and the underlying diagram indicates the changes we
are now to follow, First operation: Commencing with a portion
of gas behind the piston, at temperature ^ (that of the hot body),
pressure Px, and volume Vp we allow this to expand at constant
temperature while doing work Placing the left ,end of the
cylinder on the hot body, the expansion curve is the isothermal


Carnot's Perfect Engine.

i 2. Second operation: The expansion is continued, without supply
of heat, by placing the cylinder on the non-conducting cover; and
the adiabatic curve 2 3 is traced, the temperature falling from
rx to r2, on account of work done by the gas. Third operation:
Compressing the gas at constant temperature r2, we place the


cylinder on the cold body, to receive such heat as must be
rejected; and the curve obtained is the isothermal 3 4. Fourth
operation: Finally, place the cylinder on the nonconducting
plate and compress along the adiabatic 41; the substance is then
returned to its original condition and temperature TV

During these operations the work done by the gas is shewn
by diagram F, and that on the gas by diagram o, their difference

Efficiency of Carnot's Engine.


H being the effective work given to the engine.    Reckoning the
heat used, we have :

From i to 2 (r^). Heat expended, being work area i,

= P1V1 loge *i = ^ loge rv
From 2 to 3 (r%).   No heat expended, external work, j, being

done by abstraction of heat from the gas.
From 3 to 4 (r^). Heat rejected, as at K,

From 4 to i (r^).   No heat rejected, external work, at L, pro-
ducing internal work on the gas.

We have previously found (p. 607) the comparison of tempera-
ture in terms of r, during adiabatic expansion or compression :

from which may be deduced:


Referring to Fig. 617, expansion from 2 to 3 and compression
from 4 to i are between the same temperatures, so the ratio of
adiabatic expansion equals that of adiabatic compression : r2 = r±

And as, ~3 - ™4


V2 V4

i      »»)             »q

and — = -
v,      v.

vl                   ,                                     vl

Or the ratio of isothermal expansion equals that of isothermal
compression: r^ = r^ = r, say.

Resuming; when the cycle is complete no internal work has
been done—all is external work:

. . External work = Heat expended - Heat rejected

=  CT   lOe   ^   - *r   lOge >     -  (T    ~ T     C

Efficiency of Engine :

Work done
Heat expended

--------17ET75-------"^   & &' 7<58' 883' 887'

i v.     &«   /         ------1—      934, and 966.)

It will be easily seen that for the highest efficiency, r2 must be
nothing, or the condenser must have a temperature of * absolute


Reversed Action and Second Law.

zero/ a condition practically unattainable, and all the heat in the
working substance can never be utilised. The energy obtainable is
only that between the available temperatures, and the difference of
T-J and r2 should therefore be as large as is practically, possible.

Reversed Action occurs, as previously suggested, when
expansion takes place along i 4, 4 3, and compression along 3 2,
2 i, the operations being entirely the reverse of those just con-
sidered. External work is done on instead of by the gas, and
heat is taken from the cold body and rejected into the hot body. No
better practical example of a reversed cycle can be given than
that of an air-compressing engine as at Fig. 562, p. 546.

Let it be possible to have an engine (No. 2) of equal power
but higher efficiency than Carnot's (No. i); and let No. 2 drive
No. i in reverse order. Then No. 2, taking its heat from the hot
body and rejecting into the cold body, and giving all its externaf
work towards driving No. i, the latter is thus made to take heat
from the cold body, which, together with the work received, it
delivers into the hot body. No external work being left over, the
contrivance is self-acting.

;. Let H2 be the*heat taken from the hot body by No. 2, and h^
that rejected into the cold body; Hj the heat rejected into the
hot body by No. i, and h^ that taken from the cold body. Power
being equal,

(Reversed) H!-A! = H2-/&2 (Direct) . . . (a)

Efficiency of No. i = 5tZ*i       Efficiency of No. 2 - ^^



is to be greater than ^
6                    %

, by (a), the numerators are equal,
H2 must be less than Hx.

(See pp. 770, 773,
883, and 1132.)

The heat taken from is therefore less than that given to the
hot body, and by a self-acting process heat is being taken from the
cold and delivered to the hot body, which is impossible by the

Second Law of Thermodynamics.—Heat cannot pass
from a cold to a hot body without external aid* This is the
resojt of experience, the tendency being always to equalisation

<»|                                            M'/

|f|,«f**«f    4R|t    4**4fJ                                                      >f-

'•^/J*                         rf*lfi  Dttf            4|           %«l*^|f    *yf   *|f ft              f jj

*t                          *N                 r-*^*^4fl8   |»^fef*|                |

t *         *? $         ,

fv^Kf^     |p   I*    *"*      tJ^fl   "^          Vf*«4?^f              |f    "'tof rff                       P|

Vf^*|»'   4f^«    ^•w'*^f    »* *S/1# 4* ^ J»   *#*   # f*f f % JJHil j(    J|^#|#^ |     ^|
*t|   J*  * ^'^jl   r',   **   '                   /#^^|*|JS||           >v  l!/*"*»^fXfJ

^WM|f ^  if|«*^    M      ^      |»4Nf4|#?W|#,*»^*                   I    '4or      it     "I

l^^^fi*^ i »»t   4f^f«^   t#tii«*   '^tty               »*                     ji   ^rav^l     *

»iy             «f ^ J* *>

*^l       f »i*'t*I     ^' '*     **'•< !§€«<*«*!«*#>    jtff t£     '/|     <<|.1&Mjf    |  *-IS^^ ll»f ft

fn .3    »v    «* x i#< nf      *vi*   %         ^»i ^Itf-w'ti ,^|    l»l^0*^/      ;
|i / f      i     < «, v t  f         * <    , »f i  * > | * <i»»'        /    i>    |*V    \ i J |f  i J*   js «i   * "f ^     S       i* #• <f J | T,      f

*ii                     |i^

I*     "      *         ;>«',*<„ ^i    '   '''   * * "•"* - *    ft*' li- *   I    >  5 *• ft *>jli.J»i

%   >krt<t    •            /^*-   ,     u^9     •      .,                •«!    ;,      ***     . -fi%« >   -if nl

i, .   is<          • *           *     ,    t*,.     |jc   , ?,       •,«,     *  "   ''r*   -i t ^   .,«    , i i»^u*i,*


Initial Condensation.

8.  The range of working temperature is small in comparison with
the temperatures themselves : ^ being fixed to prevent burning of
cylinder oils and packing, and r2 by the cold well temperature.

9.  Heat is lost by radiation.

10.  The substance is lost by leakage, taking heat with it.

n. Wherever imperfectly-resisted expansion occurs, reversibility
is impaired : e.g., the 'drop' into receiver in a compound engine.

12.  Various small losses, shewn on indicator diagram : e.g.,
wire drawing, &c.

13.  Work is lost in (a) the * solid' friction of the engine parts,
(b) the fluid friction of the passing steam.    (See App. HI., p. 935.)

Initial" Condensation and Re-evaporation.—When
hot saturated steam enters a cylinder cooled to exhaust tempera-
ture, an ' initial condensation' occurs, which is not immediately
apparent on the pressure diagrams. After cut-off, further con-
densation lowers the expansion curve, as shewn dotted at AB,
Fig. 618. But cylinder and steam becoming more equal in tem-
perature, the latent heat, liberated during liquefaction, is permitted
to raise the curve, as at B c, by causing a certain re-evaporation.
The first loss is, however, very great, and by no means made up
by the second gain, so there is always a quantity of water rejected
at release, some of which evaporates during exhaust and creates a
back pressure. These losses may be mitigated (i) by applying
clothing in quick running engines, and thus securing approximate
adiabatic expansion, (2) by adopting a steam jacket for engines
of a slower type, where there is time for the heat to be taken up,
or (3) by superheating the steam before admission, and partially
removing the first cause. The jacket both assists re-evaporation
at an earlier, and consequently more available, portion of the
stroke, and prevents to some extent initial condensation : the
experimental gain being stated at from 10 to 20 per cent. Lique-
faction in the jacket is not so detrimental, but in the cylinder the
water acts as a conductor from the steapito the metal. Live steam
should always be used for the jacket, and efficient drainage applied.

Theory of Compounding^—'Another way of decreasing
liquefaction is to divide the work among 2, 3^ or 4 cylinders; and,
if great differences of temperature be employed, no other course
is possible. Thus we arrive at the Compound, Triple, or Qn&d-



J^t /ijHfi^l f^| ^t         €

J»*jfA*              irtf


v 41   *  f                        '•*•**

'tj»//'ff ^tf| frif|i^n **' i ! I* |'»jtY *



m**         *                      *


vl •

pt *m               *t§ «*ft*                   ft |?t14 It

*«| ** **           t* |» |* 4 |4f4 tit

*•§                  Jif                    jftifi   Iff,   &                           /*||   *f|  ff|

#jfy |tf        ^4         |^f                    ^*|$ *>t f 11

^1               w%                                    f                 '%          ,i  ft                 Hi

m                                                 ei»        r*        4 |t

^u*                         t  if,                                           fit

%*  /   fe                      j^|                                                  M||   |^               f«|W   flf

|P*                        ^4li</|ff^    p«|   I    %J^r(

^'                                              rf'^|    i^^«

>&1                      ^                i^    *^'^*rff  |f|l

*& ||       |»/^^4i%  igjsrMf^                ^ >i* ^**   ^


|*^ i   If

r   4A||            tff


%   ^tif
fff^J   Jftfi |<»
*  **^|i
,|ri |l


f    '    Ci',-,a ;*iv   k;«"V,   M   ivnfi.*t«*'

*•              * -

atr   provide*!,
Mali*:  two



,»!*     1

To mr itir until atot, Sif«t Sit

tin* to
*          imii<ator


**                  iliffii ribi*lhi»tli»i^r»isii«    In iilf^iii^twt-r ii'*o «n'J ?)-• < ^« «

*^«t after win* It ttu*{u]K*i         In1 rriii^iril,     I \ar '1***^   * >,\<^ *
wtit% ami ihr |MJ«-I                      ,r. * -wujiw/    ».V/ ff  ^%^


of                                                                        j ^»f,»

i>., one                                        a                  4ti<| iu%

l*i i he                       if in fin*                          <4

ii»* in I*ig» lt^.1 . ri % H llic                 MI /rt«* lit^, *i >$

» I*  till*                               «lf   15 Jl^,                   |i %  I*   iti**  nf*j l«*

ii I» lilt! ilrar^tlff                                                     lit

nf ii A,

Tin* < if»ar«iii«*              fir%l            an«l ftr  j?i«'".!?<   f.*, ••*

1¥J            tin* }ti%l«ili tl}iil*r^  |ti ri,                             S'   * ut  "'ft,   ' *|  "'  f*   •

i* lii;twrrii f* iiiiil tit uliMur |»* * %Im ,u  *i  n, MI  '   <"
tr> |             tin*              ffi«tffj%  Im^i, ^   w   . », i < «

in fh«« i oi4»l« ir^'i ' ?^ I , »' v.    f

by A j,                                               Urt*?* )v i *i4 *   ./, *, t

in               iiji to t                                     fc» ,n 4 O}» *. ** ^*4,* ^

|ti i,            if i<          til- m<«*Hi*f *>* !?ft% *ji« A t^,

t*|J ittr            lirf«»jf 44rtiif^ri »i 't^^t i/i

tttr                                   m«"; 1^* i

the                                        >*if  »1<f«f   M

IFw                ii^ , (r/^/f i%                  *»  *,  flu  *('-

un               «f                     lit *^^4f*itv ^; ^ f

It                                      tin              ff^irsUsvf

fllii                i% Irtf, tliir to          < I          **ft 1ir

t}^» rfttf <4

jfr           it   i, »»$4 *

l»f the                    on ffJ* It %ifif ^if                             ir,>

A                             » <at t%                     |l«

at tilt*             tii"                   jn4                              it,

«n                                      l*y                the                       /%»           *f A«T

Diagram Examples.


little compression would give diagrams c or D respectively, and a
shaky diagram, like E, would be produced by an indicator with
too light a spring or too heavy a piston. Diagram A shews



Indicated Horse Power.

serious cylinder condensation. The upper and lower diagrams at j
are from the top and bottom of the piston respectively in a
Cornish single-acting pumping engine : and M shews the varying
diagrams obtained from a locomotive, (i) when starting, next (2),
and lastly (3), as the valve gear is linked up from the reversing

From the Indicator diagram we may therefore deduce :

1.  The points of admission, cut off, release, compression, &c.

2.  Comparison of cylinder with boiler pressure.

3.  The wire drawing in steam and exhaust passages.

4.  The back pressure.

5.  The condensation, re-evaporation, and relative dryness.j

6.  The indicated horse power from the diagram area.

Calculation of Indicated Horse Power, or that shewn
upon the indicator diagram, and representing the work given to
the piston by the steam or gas.* Three pairs of diagrams, in
Fig. 624, are taken from the respective cylinders of a triple-
expansion engine; and are copied from the Hons. Engineering
Exam. 1887. The mean effective pressure per square inch (p) must
first be found, so the diagrams are divided into 10 parts by
equidistant vertical lines. Knowing the scale of the indicator
spring, the pressure may be measured at the middle of each
division, within the enclosed curve; these figures representing the
effective pressures. Notice that at A and F, Fig. 623, the loop
encloses minus effective pressure; every measurement must there
be treated as minus, and only added to the other plus measure-
ments algebraically. Adding the 10 measured parts, and dividing
by 10 gives mean effective pressure for each diagram; the mean
of the pair being then found by adding them and dividing
by 2.

Multiplying (p) by piston area (a) gives total mean pressure,
and this" again by stroke in feet (L) gives work in foot pounds per
stroke. Further multiplying by number of strokes per minute (N)
gives work per minute, and the whole divided by 33,000, or one

* Brake  horse   power   is   found by dynamometer,  as at p. 575,  and
.echanicaLefficiency of engine »















Advantages of Compounding.                   621

horse power per minute, will represent 4the indicated horse power
of the engine, the formula|becoming

T  j-                            /         •   x     /L0 N

Indicated horse power (per mm.) = --------

Taking the high pressure cylinder in Fig. 624, the addition of

the pressures on the left diagram, 73, 103, &c. = 679*5, anc^ ^le
mean pressure = 67*95. The right diagram similarly has a mean
pressure of 59*5, the final mean pressure becoming 67*95 + 59'5
-r- 2 = 6372. Area of cylinder is 10 x. 10 x 22 -=-7 =* 314* 16,
stroke is 3 feet, and number per minute 63 x 2 — 126. We
have then:

T      '_  .              ...         6372 x 3 x 314*16 x 126

LH. P. in H. P. cylinder = -£J------------±-2.------------ = 229*3.

33,000                    --------

In the intermediate cylinder, mean pressure on the left is 23-9
and that on the right 22, the final mean being 22*95. Area of
piston = 855*3; stroke and*revolutions as before.

___._.              ...         22*95 x 3 x 855*3 x 126

I. H. P. in i. P. cylinder = —™------------H-*---------= 224-44.

J                             33,000                    —122-

Mean pressuFes on left and right respectively in low pressure
cylinder are 9*5 and 7*65, and the|mean of these is 8*57. Then,

T TT _   .              ...         8*57 x 3 x 2290*22 x 126

I. H.P. in L. P. cylinder = —il—j2-------2——-------- = 224*41

}                             33,000                   ----ail-

Advantages of Single, Double, and Triple Stage
Expansion.—The advantage {of expanding steam in a single
cylinder, instead of using full (pressure to the end of the stroke,
was demonstrated by Watt in 1782, and can be understood from
Fig. 625. E G is the stroke, and FN that portion during which full
steam is used, the rest of the stroke, N G, being completed by the
pressure of the expanding steam. From what we know of the
work diagram, SBcu will shew work performed by the 'full'
stearn, without condensation, UCKT that by expansion without
condensation, and HSTJ that produced by the use of a condenser
Not only then do we obtain additional work by condensing, but
we are also enabled with the assistance of high pressure to
introduce an earlier cut off and higher rate of expansion: thus
using a less weight of live steam.


J* »}*' rti l-f *l #llfy / ^* ^«*l Jf*

I   J|,n}>»     '4^|t uJ^M    ^rtf4             |'«Jff^>^*   4f    |tff?|    i|t

|*4t     '4    '*    *^t|»^^|
,                                                |*»

. »f

nr 'I    I *   *

624                     Combination of Diagrams.

primarily the base line should be volumetric, so in representing
these diagrams to the same scale, the bases must be altered to
suit the volume of each cylinder respectively, and one pressure
scale be used throughout; we shall then see at a glance the
comparative work performed in each cylinder, and shall further
be able to judge how nearly the total diagram corresponds with
what should take place were the whole expansion to occur in one
cylinder under theoretically good conditions.

Strokes being equal, the area, or diameter squared, will
represent cylinder volume. The squares of the diameters are as
4 . 10*89 : 29*2. Taking clearance at \ cylinder volume for the
H. P., -fa for the I. P., and -^ for the L. P., they are represented
by -5, 1*1, and 2*65. In the large diagram, set up at MA a scale
of absolute pressures per sq. in., and measure volumes along OK
to any convenient scale. Thus the dotted rectangles CE, ZH,
and QU are obtained, in which the indicator diagrams are to be
inserted. Divide DE, GH, and JK, each into 10 parts, and erect
vertical lines, upon which pressures are to be placed, as taken from
corresponding lines on the diagrams w, x, and Y, being careful to
set them up to absolute scale; and the shaded curves are obtained.

Next mark point of cut-off B, from which to draw the satura-
tion curve. The latter being always shewn in terms of specific
volume (see Fig. 608), divide AB into 2*7 parts, or the volume of
one pound weight of steam at 165 Ibs. absolute pressure. The
method of division is shewn at ML : an inclined line is drawn and
2*7 divisions to any scale placed upon it; then parallel lines to ML
will divide the latter proportionately. The volume "41 cub. ft.
has thus been found, which being crossed by '35 Ibs. sq. in.
minus, gives the new origin for the curve BSR, to be drawn as a
hyperbola in the usual graphic manner (Fig. 620). A second
curve CTU may be traced by dividing AC into 27 parts andr
proceeding as before, the origin being then much nearer o.

By stepping the cut-off ML into the whole volume MK, the
number of total expansions 13*84 is found, the pound weight of
steam occupying at the end of the low-pressure stroke a volume
of 13*84 x 2 7 or 37*4 cub. ft. The shaded areas, then, further
represent the work done by one Ib. weight of steam, if the base
lines be specific volumes, and the pressures taken from the

/. H. P. from Ideal Diagram.


pressure scale, but multiplied by 144 to obtain pounds per sq. ft.
The gaps between areas and saturation curves show work lost, but
while there is a loss on the side s, there is a gain on side T. A
much greater loss occurs from initial condensation, which is not
here shewn, and the total 'missing quantity7 can only be dis-
covered by some construction such as is given at p. 764.
Also it must be understood that the saturation curves cannot be
exactly followed except where good steam jackets are adopted;
the curve should otherwise be nearer Rankine's adiabatic
pyV = C, which falls slightly below the saturation curve.

Calculation of Work and Horse Power from
Theoretical Indicator Diagram.—It is sometimes con-
venient to make rough preliminary calculations from a simple


hyperbolic diagram as in Fig. 625, where various losses at the
corners, caused by release, wire-drawing, cut off, &c., are neglected.
Then             Area of BN^/'

Area of CRN G - # V log* *- - f{ (/' + c) Hoge j
Mean effective pressure /« = {(areas BN -h CKNG) -r- L} — p\>


Horse Power

(See Appendices I. and III.,
33,000         fp. 111 and 931.)


J *

]n^fp                ^ /M>- ****

H * ,/n,        **
-if                 , f ,.

vrn f


***>*                                 **,JM| .ftp tit                         tufft

III Jfj         |1

j»^r*            j«   if           ** NVf* '* ,/} I11*1*  *|f*M*t »

, j.«if r^i ^ i' r^ * /i                  tit

jo           u;

V   ** j»    '

4 "IIsI ~


M|I   If ^                                                «II{1                                 «f|

JOJ                         JO

	t              [

ft    f    *>M,I
	1  ' ^^»f       liti
	K^t.i ;

|w| f          i 1                 1
	'              f    '     t'rti
	'    K1^ t   '         f

ft          tt-r*
	' $Lt       t^^i
	• i> i"
	f 10,1  !   S£ r

|n|  f             ^1,11^
	>,t         6^f|
	*?t6            Si?

III «r      H '*   ,    N^
	!«* i     t^/^
	S4 t*
	"itM        Sr r

III*       •. *       1 1 ^
	«       1 ' > s<
	1 ^ti

t f^ t      1 f #*       '» ^»
	!. i **      «ii t ,
	•   r*sS        Si

;'' * t           ^        i £*>
	\ ^      *i>J
	,   Mil*          S

' ,« *       ' f ,        r),,
	> f »i       if!
	ii « . 1
	|   icr        Sr

* i /            >         t^>«
	?j       ? s f
	\ t*
	I;                 U

Horse Power for Steam used.                  627

Such a method of reckoning horse power is convenient when
deciding boiler capacity and heating surface.    Then :

Steam per minute )      (I. H. P.) 33,000 + 144^^: AN
in cubic ft.

Steam per minute \              ,. I. H. P.      cAN } _  _

-------~~r;--------- I = 229-16               -i-----— L Df Lq

in Ibs.           j          *         prN         rV   )         *

where V is specific volume at the higher temperature, Df the
diagram factor (p. 772), and Lq the liquefaction factor (See
Appendix //, p. 884). If steam per brake horse power be desired
the value B. H. P. —77 may be inserted instead of I. H. P., where
r\ is the mechanical efficiency. Willans' important law connecting
steam consumption and H. P. is given in Appendix II., p. 892.

In the above formulae p is mean effective pressure per square
in. At p. 625, this quantity is estimated in terms of initial
pressure. If then it be required to know the volume of steam
used, in terms of the initial pressure, it is only necessary to
substitute the value at p. 625 for/. (See Appendix III., p. 931.)

General idea of the various forms of Steam Engine.
—The steam engine is a prime mover designed for converting
heat into work by allowing steam to expand behind a working
piston. Sometimes the work need only be of a reciprocating
nature; while in other cases, and this by far the greater number,
rotative motion is required, and the crank and connecting rod, or
some similar appliance is then employed, as fully set out at
pp. 486 to 496. Sometimes also a rotative shaft is introduced,
with a fly-wheel to assist in maintaining regular reciprocating
motion, where that only is needed, or perhaps to work the valves,

The Beam Engine^ though almost obsolete, has served and is
serving much useful purpose, and a few of its applications will
therefore be described. In Fig. 626, A is a Cornish pumping-
engine, a being the cylinder, e the working beam, and f the
pump-rod passing down the pit-shaft Steam is that known as
' low pressure,' having only a few pounds7 pressure above the
atmosphere; and there are three drop valves, ft, c, d, for its
distribution, called respectively the steam, equilibrium, and
exhaust valves. The last passes the steam into the condenser g,

628                    Various Slow-speed Engines.

where a vacuum is formed and maintained by the action of the
air pump h. Fig. 608 shews that water under low pressure (as
in a condenser) will boil and form vapour at a low temperature ;
and the air pump has to remove this vapour as far as possible,
as well as the condensation water. Even then there is always a
back pressure of 3 or 4 Ibs. per sq. in. When the piston
descends, valves b and d are open and c closed, there then being
boiler steam at top and a vacuum below ; during the upstroke,
b and d are closed and c is open, which places the piston in
equilibrium," when the pump rods raise it by their weight. The
parallel motion (Watt's) is explained at p. 499 ; but in A, Fig. 626,
one radius link is formed by the portion t k of the beam, and a
parallelogram then connected to the middle link kb^ so that the
valve and piston rods move on parallel lines.

A rotative beam engine is shewn at B. It differs from A in
having the crank and connecting rod instead of pump rod, and
four drop valves instead of three, the reason being that each end
of the cylinder must now be connectable with boiler or condenser
at will, and must therefore have a steam and exhaust valve. The
method of distribution is given in Fig. 629, where the left pipe
admits live steam to either end of cylinder, and the right pipe
similarly removes the .exhaust steam, whenever the proper valves
are lifted.

A direct-acting pumping engine like that at c may have a
beam solely for actuating the valves and air pump, though it also
serves to guide the piston-rod. The straight line motion is
Scott-Russell's (see p. 486), sometimes 'called ' grasshopper'
gear. A beam blowing engine is shewn at D, a being the steam
cylinder and b the blowing cylinder, the latter having inlet valves
dd> and outlet valves <?<?, for both ends, so that the issuing air
may pass continuously to the blast furnace or other place of use.
The fly-wheel is introduced to steady the motion,

E is a compound beaju engine. The high-pressure cylinder a
is placed nearest the beam trunnion, and the low-pressure cylinder
further outward. The valves are not shewn, but are so arranged
that, when the steam has done its work in the H.P. cylinder, it is
allowed to expand into the L.P. cylinder before passing to, the

630                Various Medium-speed Engines.

The side lever marine engine F, the first form considerably
adopted on steamboats, was but a beam engine doubled upon
itself so as to save room, a is the paddle-shaft, b the steam
cylinder, c the beam or ' side lever/ and d the air purnp.

The Direct-acting Engine is shewn in various forms in
diagrams G to R, Fig. 626. G is a horizontal factory engine, with
condenser a behind a cylinder <£, so that the air pump may be
worked in a simple manner by projecting the piston-rod back-
ward. By dispensing with the beam very considerable friction at
the trunnion bearing is avoided, caused as such friction was by
both load and resistance, or double the piston load. In the
horizontal engine there is, however, some additional frictional
loss, due to weight of parts and thrust of connecting-rod, while in
the vertical engine, although the former is eliminated, the latter
still remains.

The diagonal paddle engine at H, like other marine engines,
is designed to save room. Whenever paddle propulsion is em-
ployed, these engines are now chosen for the purpose. The
condenser and air pump are placed within the (triangle.7 j is a
form of factory engine seldom employed, but given as an example
of,a vertical engine with cylinder at bottom and crank overhead;
the slide valve replaces the four drop valves of Fig. 629, being
worked by eccentric, from the crank shaft.

Two other paddle engines are shewn at -Q and M. Q is the
oscillating engine, ekceedingly simple so far as the main mechanism
is concerned, dispensing with a connecting rod; but the valve gear
is more complicated than with fixed cylinders. The steeple
engine (M) was introduced to save head room in shallow boats.
Two piston rods are employed, and the paddle shaft is placed
between crosshead and cylinder; the connecting rod is said to
be '.returned.' The principal objections to this design are the
difficulty of staying the slide bars, and of keeping two parallel
glands steam tight.

The Penn trunk engine (N) and Maudslay return-connecting
rod engine (p) are examples of early screw engines. Being both
placed athwart the ship, they must be shortened in length as
much as possible. Penn got rid of piston rod length by using a
trunk piston and driving the air pump by a rod connected directly

;                                        Various Medium-speed Engines.                631

to the latter.     The practical objections were  the difficulty  of
packing the necessarily large glands, and of getting at the trunk -
pin;  but a more serious objection was  the increased  cooling
I                   surface.     Maudslay's engine was essentially the steeple  engine

;                   laid horizontally, the air pump being worked from a projection

on one of the piston rods.    The packing of the parallel glands
was the only difficulty.

The modern marine engine is always either compound, triple,

or quadruple in design, the two-cylinder compound being shewn

at L, which also serves to explain the triple or quadruple.    The

type is  known  as  the 'vertical inverted/  or 'steam hammer/

and is merely a direct-acting vertical engine with cylinder above

i                   and crank below, to give sufficient propeller immersion with direct

driving.    The slide valves are driven by eccentrics as at j, and

the air pump by a rocking lever.    The surface condenser is cast

with the standards, on one side, and the exhaust steam sometimes

passes through one of the standards;   but the method  is  not

advised by some engineers, because of irregular alignment caused

:                    by expansion.    When the triple engine is adopted, the valves are '-

either placed between the cylinders, or as at R, on one side.    In

the latter case the valve gear must be somewhat altered,    a, l>,

I                    and c are the cylinders seen in plan, and dy e, f the respective

;                    valves:  in this example  of piston form.     The passage of the

>                    steam will be understood from the sketch, entering first through

t                   d to #, then through e to £, through /to c, and finally out to the


High-speed Engines are a class of engine, usually of
small proportions, making 500 revolutions per minute or more.
A few principal examples are given at Fig. 627. A and B are
types of the rotary engine, much in favour with inventors about
the year 1870 and previously, but now practically discarded. A may-
be called the f annular' and B the ' eccentric' type, a sliding
* abutment' a being required in each case to receive the re-
actionary pressure. There were difficulties in these engines
regarding packing and expansive working. Willans' side-by-side
three-cylinder engine c, and Brotherhood's three-cylinder engine D,
dispense with valve gear. At c the piston rods a, b, c, act as
valves, each admitting or cutting off steam to the next high-


Various High-speed Engines.

pressure cylinder in order. The high-pressure pistons further
act as valves for similarly distributing steam to the low-pressure
cylinders. Engine D has a special valve of annular form, through
which the steam passes both to and from the cylinders, as shewn
by arrows, c and D are single-acting engines, so far as each



cylinder is concerned, the steam pressure being felt only on one
side of the piston ; but, taking the three cylinders together, there
is an impulse every third of a revolution, instead of every half
revolution, as in ordinary single-cylinder double-acting engines.
In both engines it is only necessary to tura on steam to start in
any position, while if reversal is required, an extra four-way plug-
cock, called a reversing valve, is interposed, whose duty is to

Various High-speed Engines.                    633

change the order of the passages, making the steam the exhaust
passages and vice versd.

The Tower spherical engine, E, and the Fielding engine, F, are
kinematically based on Hooke's joint (Fig. 475, A). In the former
two revolving bodies, a and £, are hinged on opposite sides of a
central disc or ' wobbling' piston c, the hinges being at right
angles to each other. Within the hollow sphere are four divisions,
i, 2, 3, and 4, the last shewn closed. As: the bodies a and b
rotate, and the disc c wobbles, the divisions will in turn open and
close; and it follows, conversely, that when steam is admitted to
these chambers consecutively, the said movements of the disc
and bodies will be imitated, and the shaft d rotated. To effect
this, steam is admitted on one side of the supporting web <r,
passed through proper ports to the four divisions in correct order,
and exhausted on the opposite side of e. The Fielding engine
works similarly, the practical difference being that four curved
cylinders are employed, instead of quadiri-spherical chambers, corre-
sponding pistons being formed on the central disc. A larger obtuse
angle between the inclined axes probably reduces the frictional loss.

The Westinghouse engine, G, is a type of many modern high-
speed engines, two single-acting pistons forming the equivalent
of one double-acting engine, A piston valve distributes the
steam, and the alignment of piston and crank should be noticed.
The down-stroke only being of importance, the cylinder centre-
line splits the crank radius instead of the crank circle; the con-
necting-rod's angular vibration on down-stroke is therefore halved,
and a much shorter rod may be employed, securing compactness.
During the up-stroke the rod is at a bad angle, but that is of no
consequence. The Newall engine, shewn in section at H, is
exceedingly interesting, through dispensing with so many working
parts; in fact, greater simplicity with efficiency could scarcely be
conceived. There are two sets of rings on the trunk piston,
between which are slotted holes for the passage of steam. The
distribution is effected by enlarging the trunk pin or connecting-
rod end into a hollow valve, with a partition; and ports are so
arranged that steam is admitted to, or exhausted from, the back
of the trunk, at correct times, merely by the vibration of the
connecting rod. (See J>j>. 893, 966, and 1138.)


lff> i?it * Mjf                              nn

?JM   *T<   »%fl   l«||                    **y**M          |«^                           i                ft

>***t f         4JHM*'?'*               «t                                 Jtft              111 %p

"*** Jx-               45||   'tl^i    ff||   III               %l                                        Jtff f^

*f»f|  fo   ifctv/j   jutf^   i/*              j|*|                     y    j,%|i'%

j^pir|it         |##                |f

^iri                                    If

|       »aJi!tJfli i             »»|                  J|t

(f    i$    t^*-*   ¥  '*V   i^^e   '<i*i>t

^4    ^

jflfain,  |f* KW«|,|  J|||l^f4^ til1                       ,W| f*| it

.,,^   *a^   J..K  J. t   ;-/'^ p                *|

l   t.Mj*-',14   «   ^tT4& in                             t|{        |

r         j*;*'»^   |*-!6.ri|t-*              **                                ^»t|l                          ">*<

.% .< M   1"  | JiM* > 7*J*Mr^                                      4f(i                          fpt^'i

>   ,     j»j^*v{ ii*i* ^* 4i|          **|           ^i * |.tjuj w*^**1^ ***{! ',1^*114

;       i* -^i jv                                 tiip         *

;   y i?   •.   i|«»l»*| *| *|r* Jfjs                                          ^|>irw

I    ^ ^'*'fcjkr «!-*'{ |                                                 J»{              f,|   |4»   ^'li   I**   %t

• t| /* |f**l% Jt|§                                     !*•»                          *4 t *^ *

|   #.|   j%f 41              |l                         up* 4t|| 4#| f.*1>Jf  l^ttfp*!   |f    .f*   *

t *^'*'                 **»|i |«                        vrtp t|\n4 '*t •vjir\«>.%t'i|>tik*   i  i»n|j^

»r   i                 Jifi           *|




^^                                ^^—,^,-^17^^'''^*''*^^

<|                     f»

Relation of Crank and Eccentric.


and x as INSIDE OR EXHAUST LAP, forming an additional width
to the valve face, in line with valve spindle, on the steam or exhaust
edges of the valve respectively, for the purpose of giving early cut-off
to steam or exhaust. By adding steam lap the width of opening
is decreased, which is, however, compensated by giving increased
travel to the valve. Inside lap is rarely necessary, the alterations
in valve position caused by introducing stearn lap usually giving a
sufficiently early cut-off to exhaust (compression point). Various
interesting points are raised by altering the proportions of the
slide valve, which will be fully investigated later. (See p. 772.)

Relation of Crank and Eccentric.—The commonest
valve gear is the eccentric and rod. The eccentric is merely a
convenient form of crank whose pin is so enlarged as to envelop
the shaft: it follows that the eccentricity or length of eccentric
crank must be measured from centre of eccentric sheave to
centre of shaft. This amount we shall sometimes call the throw.
While, then, the piston moves the crank, the latter in turn moves
the eccentric, and so automatically, by the slide valve, adjusts the
supply of stearn.

( Without lap.} A normal valve must of necessity be at half
stroke when the piston is at the end of its stroke—that is, when
,the crank is at a dead centre^ for then the valve should be just
opening to steam. The eccentric crank must therefore be placed
at 90° to the engine crank. Further, the direction of rotation
will be determined by the position, right or left of it, of the
eccentric. The eccentric will always lead the crank or travel
before it; for, if we endeavour to turn oppositely, "we shall only
close the steam port at the very time itf-should be opening, and so
block the supply. Therefore, in a normal valve, the eccentric must
lead the crank by 90°.

( With lap.) Let us next consider a valve having lap. Re-
ferring, again, to Fig, 634, the thin outline shews a valve with
lap, placed at mid-stroke. It then covers the steam port plus the
lap. The crank being on dead-centre F, it follows that, in order
to admit steam by port R, valve must be moved bodily to the
right, and the eccentric lead the crank by 90° + /#/, as at HX. A
little consideration will shew that strictly similar conditions obtain
with the crank on the dead-centre z.


Reversing by Loose Eccentric.

(With lap and lead.} To assist the compression steam in
preventing a knock on the crank at the end of the stroke, it is
advisable that the valve be slightly open when the crank reaches
the dead-centre. This is called -lead, and is the amount of opening
of steam port at the commencement of the stroke. When a valve,
then, is provided both with lap and lead, the eccentric must lead
the crank by 90° + lap and lead, the lap only being apparent on the
valve, while both are apparent in eccentric position.*

Reversing" Gear.—Factory engines always rotate in one
direction, and thus only require a fixed eccentric. Again,
changing eccentric from H to j, Fig. 634, will change the direction
of motion, then shewn by the dotted arrow instead of by the full
arrow. Fig. 635 gives a means of moving the eccentric to the
opposite position, when the engine is at rest. Sheave B being
firmly bolted to a fixed plate A, can, on unloosing c, be slid from
h to j and rebolted, or, still further, can be made to take any
intermediate position between h and /, giving a variation of travel
with the same lap. Such decrease of travel means earlier cut-off,
as we shall see later.

Reversing by Loose Eccentric.—But it is not always
convenient to stop the engine for any considerable period, and
Fig. 636 shews one of many methods by which a single eccentric
may be quickly changed from one position to the other, c is the
crank, having a stop D fixed symmetrically, and A the eccentric
sheave, which, being loose; on the shaft, is provided with a balance
weight E to prevent spontaneous movement. At present the
sheave has its centre at j, and while the eccentric leads ttie crank,
the crank drives the eccentric ; so, although / causes the crank to
turn round left-handed, it is at the same time pushed before the
crank by the stop D. But the sheave may be swung round to
/ or h> when starting the engine, in a manner to be described
Lifting the gab F from the valve spindle pin. disconnects eccentric
from slide valve K, when the latter may be moved by the hand
lever H. On starting, then, the left hand lifts the handle GV while
H is moved by the right hand, and thus steam may be admitted
at will to either side of the piston, according to the direction in

* The student must carefully distinguish between the two applications of
the term 'lead,' which need not, however,

#. 633.


Reversing by Link Motion.

4 K


N ,.

which the engine is to be turned. The slide valve K once opened,
G may be dropped, crank c catches up the sheave A by the stop D,
F find its way to the valve rod pin, and the gear is once more
automatic. The engine may be stopped by lifting the gab.

Reversing by Link Motion.—If two fixed eccentrics be
placed on the shafts, one for forward and one for backward move-
ment, it can be arranged to put either eccentric in gear as
required, the other remaining inactive. The gear for this purpose
is known as link motion, and, though more complicated than
loose-eccentric gear, is more easily manipulated, and is absolutely
certain in action whatever the position of the crank. In Stephen-
son's Link Motion^ Fig. 637, the eccentric rods A B are connected
to either end of a link c, curved to a radius from D. The valve
spindle F supports a die E capable of vertical movement relatively
to link c, such movement being controlled by the lifting link G.
At present the radius link is in * mid gear/ and any * plus ' move-
ment of one eccentric rod would be met by a ' minus' movement
of the other rod. If these movements were equal, the valve
would not travel at all; but, as the sheaves are not placed directly
opposite on the shaft, the plus and minus displacements do not
balance, and the valve opens-to lead.* If the reversing rod H be
moved to the right, the rocking link G will lift the radius link
until rod B is nearly level with the valve spindle, and the valve
then receives almost all the horizontal movement of the B, while
A'S motion is all but inoperative on the valve. The eccentric B
is then in ' full' gear. If H be moved to the left, the A rod is put
in gear and B is practically inoperative, j is termed the weigh-bar
shaft, and H is coupled to a hand lever on the driver's platform.!

In Gooctts Link Motion^ Fig. 638, the eccentric rods A and B
always vibrate at the same height, and radius link c rocks from a
fixed point G. But the valve rod is in two parts, one of which, K,
the intermediate valve rod, being lifted or lowered, changes also
the position of the die E. In the figure, K is shewn in direct
connection with the rod A, while B'S vibration has no effect on the
valve. When K is at its lowest position, rod B is in gear and A is
noperative; link c has it curves struck from r>. It should also

* Larger lead than that ia full gear.

t In large marine engines it is usual to reverse by steam power.

J     •'>

1*.  /, '>.'*','*

*1 ?n i * 1     , f m'^  .


mjf1.*!'    At I    I       <''rv»/**r

f **«*'*' tr* i i<r \ *  &i vt

Hf .ils |^»^l  f •rd    -7^  it f
fir    .^f'lf*1!   *}M*.   <   ',       *»
*| f^r   l/^ I?* ti i* i *  « f ti

Ill      ,   '   '.                     t

,-.,',            1    ',*          I   " I

'   ,/     i             I                   I fc         />,               '.      1   4 f

I r -    \ ' ,     /; ^              %     ,     *           t

A  *             f      U   /

" 5.

t I1 %     t <

»**     *l

*   <     4*   *Mlv  •*"'

<         t 1



»'                                     fm

tft C§*                  ^.l^ftf*

'lr*f     **„«'   ' l-Pff"'1'   >**#*• ill *4*/'^ ^ |            **                          t\^n                     ftt

I    j^f'jf    J>*^M                "^(t. (f|    **    ' if Apt I #     ^f||   «v^                        |               (4f

4    /*lsvf*4|   I'. .-^    'ff t'l    !*^|*J«,i»   jf«*   •»*'»   *   ***

jf'4|4S/^#^| jt'fg    "#    t <l   '<}   4/>   f  f   <7|    rf  |0 fc * fr«»j'  Ml 'Vri*^ ^|

P'*A^*T|       «'^    |^      *^^/«     ^'      ^t         /',l|'J#|*     l'|     |^#'V|*^|       |^

^s-^Jt    .i^.fl    4   |   |*#|,Ai |l^^'a   %*    *\  ft   v,^n/'(^   f|V«*    ' *"f,

§ <i   |w  il'l^f^'Jf ^**J*f   dfr'H   ^|'^Vf^*i, js   ^ ja
•'(^S.?1**^    *-'M   *"   ^*         |f ^J%l" *t ^|f  Alt ^I'fft.^'f^I    t€    ^i    4tjpft    »#> | ^   fef

4         *   %             i   ;    i' / ' ^   /"¥   ^4#J *    I   >5^4   ,^',-f   ^ ' Jr      »   J  (*   »tf   f iK##t ^

' ,^ '•'"*!.', i** • « u,    >• i / "     / , * ifi   i     MI >'/1   •*    i*» f  »^r>

N .1

Oscillating-Engine Valve Gear.

An intermediate valve rod F, may, however, be changed from
D to B, Qi^mce versa,) by the reversing lever E, so that F may move
either in the same or the reverse direction of j. When F is at B
the eccentric must lead the crank, as in Fig. 634; but when F is
at D, the eccentric must follow the crank. The intermediate rod
F, again, is only connected to the valve rod G through the lever
LM, the pin K forming a fulcrum upon which LM is rocked by the
crosshead N. The travel, LP, thus obtained, represents twice
(lap + lead), as at hj\ Fig. 640, and takes effect at the dead centre
positions. When F, therefore, is in mid gear at c, the valve opens
only to lead, but when moved to D or B, the opening is eccentric
throw minus lap, as in Fig. 634.

Valve Gear for Oscillating Engines.—The method by
which a satisfactory motion of the valve is obtained will now be
made clear by reference to Fig. 644. T and u are the valve boxes,
of which there are two, in order to keep the cylinder balanced.
Y is the cylinder and v w the trunnions, being steam and exhaust
pipes respectively, supplied with stuffing boxes. M m and N n
are the valve levers, rocking on fulcra R and s; and P Q the valve
spindles, guided at their upper end. All the parts mentioned
share in the rocking motion of the cylinder, the remainder are
either fixed to the snip or take motion only from the crank.
z z are fixed guides for sliding link L, whose slot is curved to a
radius from trunnion centre. To L is again connected, by centre-
pin F, the usual radius link GH, which is moved by eccentrics cd
through rods D E. x is the trunnion bearing.

It will be seen that the rocking of the cylinder can in nowise
affect the vertical movement of the valve levers; but any motion
given by the eccentrics to the link L is faithfully transmitted to
the valve spindles through their levers, the discs j K always lying
in the link L. On account of the introduction of the rocking
levers M and N, the eccentric motion will be reversed. The
eccentrics are therefore set to follow the crank by 90° minus lap
and lead, and the rods are said to be crossed.

Steam enters at v, and passes into the valve chests by the
belts e e, entering through port a. After giving work to the
pistons through either steam port b fr, it exhausts through the
mid port ?, and passes out through the belt f to the exhaust


Oscillating-Engine Valve Gear.



pipe w. ' Sketch g is a front view of the ports. A loose eccentric
or single fixed eccentric may replace the link motiqij,'but the
link t is always required.



-  If

ft   '

tr  %/4

*' .^ jft


'»»«*#    >t » * f     t *

^       i?       v


i *! *

•s - >


"  1

5^,    %

^     *

1  S

«t     *
*1f    ^

» It



|f             ^|              4^|                If   ^ 4

*   «*          4   *f^f    t*|

i'«jK»^     H               *»|n*l***?f   «j*fft     *%   ***|                                     4»|

*»<%»(> i* |t««^,|  j^ it'^i   || |^^t   i     *f rf-          *frf^  ^t

,4   Ji*    |4/||*    r^|   ^JH|f        |»*r'H*   ^'I'J  |^   ^»**#   ^|l

>;              *»"14fc*     /^    ^f%*s    1    ^f^cir              '*«••                                       ||

J^|/**/f   ^     fy*»ii#t S>-             #tf                                    ^Mli^ftf   ^ffe|j

f|    l.f|   "^«^t   tn***'*itHtft   *'VI            ***                   /^^xfc             j^,|

*|+*1|f^  ,#<        j/| ,?;i  |^|.*^'%   ^yr{,O*i **                  Jt*4\. jtt                !&fr >»||


**£   <;«? f    j, ^   "fit    t*


r.    t       .»'/<r     14    ?     *         \      .>;/.**<.!*      -MM     *.*5     V^fl7'*      I,'/      *<<

,0-v-   m*      if   \<wi   ;• T?,    »-^   M'»^*    *K    >(l" f

it1*'       4       T'       i „ | 'J" * *      *'\*     ' '< I   '   * f * *      ^       '        s       K |         -,'

f ,, |    ^ U        i «, , v   S       «?*)-'* ^ i'J -C J»    »         /   |    ^   t «i * *f^ '    , '   >   I   *  '      '


£ J   ~   2

^   V   %   £ „£  ^

U"      **       I*


T *

^    **     £.

%;     .      *

I     *

i i *

i c t

<* *

£ 2 I

* «

* * ^

»  S   .


% jf  ?

*  *^   -> **  ^

^ i  * «*  ^ .*

-: I *•*

**     jajfc                ^^                 «^


4        f
""^         4-


—; =»! ^ il^ l| t

^      **     ««».             *      ^       *              *       *      %

X       ^               ^|*t^^*»*

. ** ^ s •* i ^ * ^ r % ^


imt^f^ffM *
M^Vi^     /

*                 I »

*l4   yf            *,$          Off*  t

#**   ir^                                                   #f

*j*Jt   1*|*                 ^ f^l*

IPSff   'hty«*j'*<   *      jm|i      >)*    A*^^   *-V«4
ufl    ft^    ^>^Jfff?t     J^;|     >»|f*    ty                £     jt*|f

/fl»        '/ **^»    '  tp f             it-»*8  /f    ^   *>|«t    J&KWlt*


V*"1"* ^

f »   »W

* •   T / v»j    #    **^ ,11'

i    '<    4     ,jt  '£''»''   *-1


Automatic Expansion-Gear.

learn that very little economy was thus secured. A £ gridiron '
valve was adopted for the expansion valve, the ports being split
into eight or nine portions, for'reasons to be explained in the
next paragraph. Clearance is decreased as much as possible in
the back-cut-off valve, especially in Fig. 648, though it must
always be greater than in a single valve.

A Double-ported Valve, as in Fig. 647, is usually adopted for
low-pressure cylinders of marine engines. As frictional loss
depends directly on distance travelled (total pressure being
equal), it is advisable to decrease the travel as much as possible.
This may be'done by dividing the steam ports into two parts,*
as at AB; only half, travel is then required. Of course the valve
must be made somewhat larger, which increases the total pressure,
and consequently the force of friction; so a portion of the back
is often shielded or ' relieved' from pressure by the ring c D,
which lies in the annular groove E F, being kept steam tight with
the back of the valve by the springs GH, and with the groove by
the ring j.

At Fig. 648-is shewn a back-cut-off valve with double ports, the
main valve being designed to shorten the steam ports and decrease
clearance. Fig. 649 shews double port?, both for Meyer and main
valve, the arrows indicating the paths of live steam and exhaust.

Automatic Expansion-Gear.—Instead of connecting the
governor sleeve with the throttle valve, as at (a) Fig. 645, it may
be allowed to alter the travel of a back-cut-off valve, with in-
creased economy and direct action. The most common arrange-
ment is shewn generally at Figs. 271 and 272,'pp. 270 and 271,
and in detail at Fig. 261. The expansion eccentric is coupled to
the central pin of the radius link, the latter rocking on a pin at
the upper end. When the governor sleeve M, Fig. 261, rises, it
lifts, by lever u-and link K, an intermediate valve rod. Thus the
height of the governor decides the height of die in radius link,
and therefore the amount of travel on the expansion valve.
Eccentric travel remaining cojptant, when the engine speed
increases and the govew@r sleeve lifts the die nearer the link
fulcrum, the travel of the valve is decreased, cut-off is earlier,
-and less work done. This brings back the speed to the normal.

V                                 * Or more, as in gridiron valve.


Marine Governors.


If, on the contrary, a heavy load is put on the engine, the
governor revolving at a low height gives the valve the utmost
travel, securing a late cut-off.

The Shaft governor provides automatic cut-off by a very
simple and compact arrangement, especially adaptable to small
high-speed engines. The gear of the Westinghouse engine is
shewn at (£) Fig. 652, the object being to directly vary the
eccentric throw. A A is a disc fixed to the crank shaft, having
pins B B for carrying centrifugal weights E E, and pin H for
supporting the eccentric H j. The latter may rock to the right or
left on pin H by a limited amount, to be determined by the
position of weights EE, their deviation causing an alteration in the
eccentric throw. The weights are connected by the link c D, so
that their movements shall be simultaneous, and are attached to
the eccentric by link F G. If the engines then revolve at a high
speed, the weights E E fly outward and pull the eccentric sheave
to the left, decreasing throw and producing early cut-off; if the
speed decreases, the strong springs K K bring the weights towards
the centre and increase the throw. (See p. 1145.)

Marine Governors have always been difficult to devise,
and, although no perfect governor exists, the arrangement at (a)
Fig. 652 is one of the best, acting as it does on a direct
principle. The fluctuations in speed of a marine engine are
caused by the propeller either partly or entirely leaving the water
rather suddenly, thus decreasing the load. The consequent
increase of speed or £ racing' cannot be entirely obviated, but may
be considerably modified by the use of Dunlop's governor, c is
a large pipe communicating with the water near the propeller, and
D an air chamber which can be shut off from c by the screw-down
valve A, worked by hand-wheel B. F is a pipe containing only
air, the entrance of water being prevented by the baffle-plate E.
H is a diaphragm in communication with F, and L K. a rod which
partakes of the movement of H, transmitting it to the piston slidb
valve M, for admitting to or exhausting from cylinder p. The
cock L admits steam to M, and the piston rod Q R is connected to
the lever R s, which has its fulcrum at w. Finally, s T is a rod for
actuating a throttle valve in the high-pressure steam pipe of the



g.  65 O.

Corliss Gear.


If the propeller sinks below the normal, water rises in D, and,
compressing the air in F, presses on diaphragm H, lifting K L and
moving K z round fulcrum z. Valve M being opened to steam at
the bottom end, piston P is raised, thus depressing the rod s x and
opening wider the engine throttle valve. But, as s moves down,
the lever K z is turned round K as a fulcrum, and valve M is once
more placed in mid position. Suppose the propeller rises, the air
in F becomes more rare, and spring j moves L K downward,
opening M at the top, bringing QR down, and raising ST, thus,
partly closing the throttle valve. (See p. 1-144.)

Corliss Valve Gear.—Of all the ' trip' gears,* this is
probably best known. In Fig. 650 the upper diagram shews the
valve gear, the lower being a section, through the cylinder and
valve chambers. There are several advantages possessed by this
valve arrangement and gear, some being common to other trip
gears : (i) a sharp cut-off is obtained, when the 'trip ' takes place,
preventing wire-drawing; (2) an easier-working form of valve, g,
is adopted; (3) steam and exhaust parts being separate, there is
less loss by initial condensation; (4) clearance is very small ;
(5) the variable cut-off" is automatic.

The valves a& admit steam, and ee pass the exhaust, being
represented in plan at g. They are hollow cylinders having a

* Term given to rapid cut-off gears, worked by the trip of a valve lever.


Trip Gears.

large portion cut away, and are rotated by spindles to which they
are connected loosely.    The steam pipe is shewn at £, and // are
the exhaust pipes, forming the cylinder supports.     Taking the
valve gear, A is the eccentric rod,' which by a to-and-fro motion
rotates the disc or wrist-plate B ; to the latter are connected the
four valve rods, two of them at cc actuating the exhaust valves,
the other two at D D working the steam valves.    The exhaust rod
CE is directly connected to the valve lever EF, and moves it
through rather less tban 90°.    The steam-valve rod DG is more
complicated, consisting of two parts: one, DORP, attached to the
wrist plate;  the other, QNG, connected to the valve lever GJ.
These tend to separate, by reason of the force in the compressed
spring T, but are prevented by "the spring catches PP.    If, how-
ever, the latter are prised apart, spring T is released, and, pulling
j rapidly to the left, closes the steam valve.    The prising action
is obtained by the toe lever MN, which, pinned to QNG at N, rocks
on ftiferum M.   As the pin D moves from z to D, the rod D G takes
a more crosswise position relatively to the toe N, and at some
intermediate position the catches PP liberate the parts r> and o,
permitting the valve to be closed.    When D moves back to z, p p
regain their normal condition and D and G are connected.    The
position of fulcrum M determines where, between z and D, the toe
shall release part G, and this is decided by the height of governor
sleeve, the latter being connected to rod ;;/.    When the governors
rise, ;// is pulled to the left, moving M an equal amount to the
right, levers WK and WL being geared together at x.    This causes
the toe to separate PP at an earlier part of the stroke z to D, and
the converse will happen when the governors fall.    Lastly, the
dash-pot s being full of air only capable of passing out at u,
reduces the shock caused by the sudden release of the spring T,
the set screws serving to regulate the air passage, and the back
chamber v is usually connected to the condenser to ensure decision.
The Proell Valve Gear is another good trip gear.    The
lever D E, Crocking op fulcrum E, may for the present be looked
upon as rigidly contacted to the arm F F, and toes F G.    At point
E is attachejl the eccentric rod, and/a movement of D to the right
will cause the left-hand toe G, trailing along HJ, to finally slip,
when spring L closes the steam valve B.    Meantime, the right-

Proell Gear.


hand toe, which tripped on the last stroke, must be replaced
on j K, and this is attained by making the L-lever x F G free to
turn on the pin F, until G is high enough to slip into place The
dash-pots are similar to those already described, set screws M M


adjusting the compression of the springs L L. A rigid bracket s s
supports the governor gear; within it the hollow spindle TT re-
volves, and the balls, flying outward, pivot R x on x, raising the
central weight p, while lifting pins vv and the spindle w to a
higher position This rise affects the positions of the toes GO,


Zeuner Valve Diagram.

bringing them nearer together ; the reverse happening when the
governors fall, and thus is obtained automatically an early or late
cut-off respectively, A dash-pot within P damps small vibrations
on the governor, entrance or exit of air being adjusted by screw Q.
The steam valves B B, being double-beat, are balanced, besides
requiring only half the lift of a single valve.

Zeuner's Valve Diagram is a graphic and ready means
of finding the various positions 'of the engine crank where
admission, cut-off, release, compression, &c., take place with a D
slide valve, when the valve dimensions are known; or, conversely,
of finding valve dimensions when certain crank positions are
given.                                        .                    •:•'..

Imagine a valve without lap, and let CD he the eccentric
throw or radius at (i) Fig. 653, When the eccentric is at D, the
valve is closed, and, when moved to H, the opening to steam on
the left is CG. Turn G round to LJ then, C.L is steam opening for
eccentric position H. A series of points such as L may be found
and the curve c L B drawn, whose radii vector' shew gradual
opening and closing of the left-hand steam port, The left diagram
being obtained similarly, join A K, Then triangles c A K, cj F are
similar and equal, and A K c is a right angle ', the two polar curves
are circles, and while circle c B shews steam opening, circle c A
represents the opening to exhaust', together being known as curves
of position for a valve without lap.

Taking a valve having lap, both to steam and exhaust, its
position curves are those at (2) Fig. 653. For the opening either
to steam or exhaust will be that at (i) less the respective lap.
At centre c, strike arc M N with radius =? lap, while p P = exhaust
lap. Then 4 B is full opening to steam at left-hand port, eccentric
being at B ; Q is admission position, and R that of cut-off. Simi-
larly AS is full opening to exhaust at right-hand port, eccentric
being at A ; T the release, and s the compression position.

We must now translate eccentric position into crank position.
Still assuming .a right-handed rotation, we, must turn back in a
left-handed direction all the eccentric positions, through the angle
by which the eccentric leads the crank, to effect the above purpose.
This angle is 90° plus the angle of advance.* The change has

* Angle pf advance == the a$g!e whose sine = (l^p + lead) -f throw,

P&SiTlON        FOR     RLL     CONDI TtOMS       OF     VALV£


2Pffr             « 4 «»-                               |»                   %

,   /*   'f, *-<•;      ^. '   f-t*     fr    t-t  ;*>4ijfe».v|  f"

4'^/it «*  pv* * **f H   ^4r P; f/f                         ft i

|$f«f  4WW*              '   fc  f

^ /«     '    *^',  j»       *,   fr<   ^4*     y         V*             *$4i,4#*>   *   4»f

1»|'>     ^>**r/    *^^llh         A/r'   V'J'   f'     '^^f,'f*^l                '%'*•*,

$                 ".       B      /••        >

V-    '  ,        «.     '         '         /»-       ,'t         .'         f

it     v  ."*.•'          ; >es  f /     *  . •

,(S^f.     «,<   t.»                      .«      *   *.»

f  I   1         ^   ,   f  '

«*                k    f    d,:    *         * ' '.  4     ,« *      f     '

i             t    -        /        ,),*,•,,       t   t

V    ,                 '.*•..,,,'

664             Zeuner Diagram for Meyer Valve.

(4)  Given travel, or opening to steam or exhaust; also both
laps, and lead.    Strike travel circle and mark points w, Y, and x;
diameter BW being known, the steam circle is struck and BA
found;  and the rest easily completed.

(5)  Given steam opening for any particular position of crank,
position of crank at cut-off, amount o   lead, and exhaust lap.
This is answered at (i) Fig. 654.  .Draw opening i, 2:.lead i:
position of crank for that opening, 3 : and position of crank at
cut-off, 4.    Drop perpendiculars 5 and 6.    Draw 7 at 90° to 4>
and 8 at 90° to 3.    Bisect angle 5, 7, by line .9 ; and angle 8, 6,
by line 10.    Their meeting point is the centre of the- diagram,
the dark line shewing the primary circle.

(6)  Given the lead, and the  positions  of crank at cut-off,
release, and compression.    See Fig. 654, diagram (2).    Let i be
the lead, while 2, 3, 4 are the positions of crank at cut-off, release
and compression respectively.    Drop perpendicular 5 and draw
6 at 90° to 2.    Bisect 5, 6, by 7-, and 4, 3, by 8; their meeting
point being the centre of the diagram,

(7)  Given lead, maximum opening of steam port, and position
of crank at cut-off; also inside lap.    For solution see (3) Fig. 654.
Let i be the lead, i 2 the greatest steam opening, and 3 the
angle of crank at cut-off.    Drop the perpendicular 4, and erect 5.
Draw 6 at right angles to 3, cutting 4 in A.    Bisect 4, 6, by 7,
and produce at 8 to G: join 9,    Draw 10 horizontally, and with
centre A strike n; join AB by 12.     Draw  13 parallel to 12,
cutting 7 in E, which is the centre of the diagram.

Zeuner Diagram for Meyer Valve.—Concerning cut-
off point only, the real opening to steam will be due to the
relation between main and expansion valves at any moment. In
Fig. 655, let AB be the stroke of the main valve, CG its, steam
circle, and 0 the angle of advance. Also let 6^ be the angle of
advance of the expansion eccentric (nearly opposite the engine
crank), and c £ its throw* Taking position ;E, c F would be the
movement of main valve from central position, aud CD that of the
expansion valve, the difference or relative motion being DF,
Measuring this difference at GJ for several positions sux% as E,
CJTK is found, which may be proved to be a circle. To ftnd GK
directly, join HG and 'complete the parallelogram HK, fcy,parallels

Zeuner Diagrams for Link Motion.


G K, CK. Then, with certain limitations, the radii vector of CK
Mail shew closing to steam of the Meyer valve for respective
positions of engine crank.

Let the back valve be adjusted to any desired width, and R be
measured when at mid position; with radius R describe the circle
L M N o. Strike the steam lap at P R Q ; the vectors within p T s u
then shew opening to steam for the respective crank positions.
Admission is given at p by main valve, in the usual manner;
after u the opening is also controlled by the R circle, and when
the difference vector equals R, as at cs, cut-offtakes place. We
see from this diagram how decrease of R secures an early cut-off
and vice vers^ and rapidity of cut-off can be judged by decrease
towards s of the vectors of the shaded area., The exhaust circle
is governed by the main valve only. (See Appendix //.,/. 895.)

Zeuner Diagrams for Stephenson Link Motion.—-
KLnowing that decreased travel when linking-up causes earlier
cut-off, we may now examine, by Zeuner diagram, Fig. 656, the
exact result obtained. Taking the upper diagram:

Open rods. With throw as radius, strike travel circle F E, and
draw valve circle D E for full gear. Draw the link A B, represented
by full travel of die, with D A, D B as the distance between die and
sheave centres. Through point G, where D A 'and valve circle
intersect, draw EGH to meet the centre line DC in H. Then DH
is the diameter of the valve circle in mid gear, and any other
circle, as D <?, will have its diameter bounded by E H ; position <?,
between E and H, corresponding to proportionate position between
A and c The centres K L j will form a parabola, lying concave
to D and with vertex at j. Draw the lap circle ab, and ^ erect
perpendicular */Y. YE will.be the lead in full gear, and the
amount of lead will increase as the travel decreases, shewn
by the shaded figure, being dn in mid position, or equal to
the lap.

Crossed rods. In the lower diagram, the full-gear circle o P is
set out &9 before, also the link M N. The crossing point R, made
by the further rod MQ, is joined to P, when SP bounds the
diameters of valve circles. The centres of the circles now make
a parabola conyex to o, and with vertex at T. Strike the lap
circle /& and draw the perpendicular ^ x. x p is the lead in full

666                           Ideal Diagrams.

gear, and the shaded portion indicates the change in lead value.
Decreasing towards the centre, it vanishes entirely at w, where the
opening is h w. At o v the throw is equal to the lap, and there-
fore the valve does not open at all at positions on the link
corresponding to between points v and s.

Space prevents us giving proof of the above, which, while
being only approximate, is quite near enough for practical
purposes. (See p. 1146.)

Ideal Indicator Diagrams for Compound Engines.
— We examined in Fig. 622 the form of diagram we should
expect to obtain from a single cylinder, and in Fig. 624 some
actual diagrams from a three-cylinder compound. The forms in
the latter case were sufficiently clear to shew considerable
difference of character over those taken from a single-cylinder
engine. We shall investigate the ideal diagrams for two-stage
compounds, believing that a careful examination will enable the
student to carry the method to three or four-stage compounds.
To simplify matters, we shall work with numbers instead of letters.
Naturally, in building up such diagrams, the only question we ask
from time to time, is, 'What is the change of pressure with a
particular change of volume ?' Two formulae are needed to meet
all cases.

(r) When the volume increases or decreases regularly within
the same vessel :

(2) When two or more vessels, having each a particular volume,
and each containing gas at a particular pressure, are
suddenly placed in communication :

P final

+   v   -1-  v

Also, for simplicity, the hyperbola is taken to represent the
relation of pressure and volume. See Fig. 620.

I. Tandem Engine^ with one cylinder behind the other, and
both pistons on one rod. Sketching the cylinders at A, Fig. 657,
we adopt the artifice of applying a i^ovable paper strip B to



* M


«|i   Jff


f    F



ti   ' *r •


l^| |     $11

ti it

'<l   ^|

« k     ,i     Vi *i         i *» • | ,   )*    i      |
f      |                   *        •      f    ,'?       «,

? '       *     '   ^   „



Side-by-side: Late Cut-off.


id now expansion takes place simultaneously in H.P., L, P.,
d R, up to cut off in L. P.    Using formula (i):
Tl 4


•4 4-


4'34   2*1


ich brings P3 up to the general hyperbola in the L. P. cylinder,
should be noted, however, that only certain proportions between
i three vessels will do this, and the curve may be either above
below the general hyperbola, as in Fig. 659. The rest of the
/-pressure diagram, Fig, 657, is easily understood, the corn-
sssion curve being a hyperbola through PL. Intermediate points
;ween P2 and P3 can, of course, be found by calculation.
Following the H. P. diagram, the steam will now be only in
nrnunication with the receiver, and must therefore be com-
rssed in H. P., Hc, and R, from P3 to P4, by formula (i):

Being cut off to exhaust before reaching P4 we assume a con-
lient point PB, and measuring, find PB = 37, or the residual
ssure we started with.

II. Compound Engine with tranks at right angles: cut off in
JP. after half stroke (at *6). This is worked out in Fig, 658, the
alt being as follows :

( H. P. = i                      -        -

Volumes-^ L. P.  - 3$
( R.       - -45

Cut off   ...    H.P. at -45,       L, P. at '6.
After first cycle it is found that

P» » 17 "7, and PL is assumed at 18 as before,

Arriving at Pj » 62, the H. P. piston will be at the end of its

»ke, but the L. P- piston will be at mid stroke,    We therefore

ce a sudden communication with R, Lc, and half L. P., all

ch will have the same residual pressure as H, and

62 x i| 4- 177 ('45 + '3 + *2> _ _,.

*•-     •-'H+V'MT^s"    3 4

Lc   =   -3

Side-by-side: Late Cut-off.

Next the L. P. piston moves to cut-off at P3, but the cor-
responding movement of the H. P. piston is so small that it may
be neglected. Expanding regularly in all the vessels, we have :

il + -45 +

P       ---     n T • A      V     __________5_______I~_____

+ '3

=     28-6

s      °         .ij + '45 + if + "3 + ('I x 3'5)

The rest of the low-pressure diagram up to PL will be understood.
Following the H. P. diagram, P8 is compressed regularly to P4 so

" + '45

P4 = 28-6 x


. + * + *45

And now there is a sudden communication with the clearance
Lc, having a residual pressure of 18; therefore,

p . = 4*7 x (i + \ + '45) + '3 * l8 =    6
5                | + £+ -45 + -3

Then there is a gradual change of pressure, all three vessels
being in communication, but the curve is not a hyperbola, because
not only are the cylinders of different area, but the piston speed
varies considerably. At centre B strike the larger semi-circle, and
at centre A strike the smaller semi-circle, to represent respectively
the L. P. and H. P. cranks. Assume the H. P. compression
point 6 and join 6 A, then draw B 6 Nat right angles to 6 A ; also
divide the portions between 5 and 6 on each crank circle into
equal parts, and letter as shewn. Now the total volume at any
point between 5 and 6 can be found, it always being (Hc -h Lc -f R)
+ vol. in H. P. + vol. in L. P. Thus at P5, volume = (J+ -3 + -45)
+ •5+ 0=1'375. For any other position, w for example, the
volumes in H. P. and L. P. may be found by taking off both the
distances * * with dividers and measuring these by the scale FG,
We have not space to consider every point, but at P6 vol. will
clearly be "875 + *o8 + *7$75 = 1742. Then,

Intermediate points between P5 and P6 on H. P. diagram
being obtained, an arched curve is found as drawn. The H. P.
diagram is next completed by drawing a hyperbola through PG.
The L. P. curve from P5 to P6 must next be drawn. Now the


Side-by-side: Early, Cut-off.

pressures have already been obtained for these points, and it only
remains to define their volumetric position. To do this take all
the points from 5 to 6 on the larger semi-circle, and transfer them
to the left side of the circle; thus B 6 is changed to B j. Pro-
jecting these downwards we only have to set up the heights
previously found, to complete the L. P. curve from P5 to P6.
The further expansion from P6 to PK is only in L. P. cylinder and
receiver. Therefore,

'45 + *3 + "7875

P, = 28-;

'45 + *3


or we have arrived at the residual pressure assumed at first.

III. Compound Engine with cranks at right angles; cut-off in
L. P. before half stroke. Referring to Fig. 659, and taking the
following data :

f H. P. = i                    Hc - -i

Volumes < L. P.   = 3                .    Lc   = '3'

IR.    = 1-5

Cut-off ... H.P. at -3,            L. P. at -4.

PB will be found to be 30*2, while PL is 18 as before. Pi = 44
by measurement. Then the drop to P2 is much less than that in
Fig. 658 because the receiver only is opened to H. P. cylinder,

and         ^       44 x ri 4 30-5 x ix

p    __ j_f________L_~_Jl_:___......_.   —   ^6'2

Compressing in H, P. and R,
i-i + 1-5

P3 = 36-2 x

= 44-8

ri 4 1*5

A sudden expansion occurs by opening to Lc, and

•n       44'8 x (I 4 'I + i%\ + 18 x -3

p   =---------~.---------—-_—_—- = 41-4

4              j + -i + 4 + -3

Then, while L. P. crank moves from 4 to 5 on the large circle,
the H. P. crank moves through a similar arc on the smaller circle,
at right angles to it, as before.    Taking volumes at P4 and P5 we
41-4 x {-5 + -i '+'1-5 + -3} _   g
•i + -i + i-5 + -3 4 *6i     ~.3

Correction for Inertia.


Finally, expanding from P5 to PR in L. P. cylinder and re-

PR = 38 x

i'5 + '3 + '61
i'5 + '3



the  residual pressure.

While a small receiver should be adopted in Case II., a very
large one is advisable in Case III. in order to equalise the work
in the H. P. and L. P. diagrams. Of course, Case II. compels
a large gap in the combined diagram, on account of drop in
receiver and low-pressure cylinder, and the arrangement is not,
therefore, counselled. The student should compare actual
diagrams with ideal ones, and endeavour to distinguish between
Cases I. and III.

Correction of Indicator Diagram for Inertia.—The

indicator diagram, as obtained from the cylinder, does no more
than transcribe the changing pressure and volume on one or other
side of the piston. The actual pressures tending to move the
piston are not correctly shewn, at least not without a small
correction; but those transmitted to the crank, which are what
we most require to know, are considerably different, on account
of the deductions and additions required to respectively start
and stop the reciprocating parts at the beginning and end
of each stroke. We shall now examine the modifications to be
made in the indicator diagram in order to arrive at the tangential
pressure on the crank pin; and, to make the investigation as
useful as possible, shall take an actual case of a vertical engine,
where there is not only the inertia force to contend with, but
the dead weight of the moving mass. In a horizontal engine
there is no such dead weight, while in a diagonal engine the
pressure along the incline caused by the weight is the effective

Let the crank circle, j K LM, Fig. 660, have a radius of i' 9",
as measured by its own scale. Divide the circumference into,
say, 20 equal parts, and, with a connecting rod 7' 6" long, mark
corresponding positions of piston stroke from A to B. Draw the
polar curves, KU and UM, by the method given at p. 491, and
transfer the ordmates to the base A B, so as to form the velocity

674                        Correction for Inertia.

curve AXB.    Supposing the crank to revolve uniformly at eighty-
eight revolutions per minute, the velocity of crank pin,

v = 16-i ft. per sec.

that is, XY should measure 16-1.     Dividing this ordinate into
16*1 parts will give the scale of piston velocity.

Next find the acceleration curve, QTR, adopting the method
already explained at p. 492, and illustrated in Fig. 454. QT will
shew, from base AB, the rate of increase of velocity, and TR the
rate of decrease, for the top diagram, viz., when the crank moves
through JKL; but on the return stroke, from B to A, lower
diagram, RT will be acceleration and TQ retardation. The
acceleration scale will not' be the same as the velocity scale,

but must be compressed in the ratio - as  explained on p. 492.
In other words:

Reading on acceleration scale

= reading on velocity scale  x —,

Produce x horizontally to Q.    Then

_     ,.                 i6'i x 16-1          0

Reading A Q = ------;-------- = 148

By dividing AQ into 148 parts, an acceleration scale is there-
ore formed. (Seepp. 932 and 1107.)

Now the force required to produce a given acceleration in a
given mass (p. 473) is -f\ that is, the inertia force is propor-
tional to the acceleration. The weight of moving parts in this
engine is 8030 Ibs., and the inertia force at any moment,

_      wf     8030                  .          ,.

F = —- = —— x acceleration reading.
g       32'2

The acceleration curve may then be transformed into a curve
of inertia pressure (total) by multiplying by the above fraction or
by 8030 ~ 32*2 = 249'4, that is, the distance AQ must be divided
into 148 x 249*4 = 36,911 parts. This has been done along BP.

From the total pressure scale take 8030 Ibs., with dividers,
and move the curve QR down by that amount, to *f?, thus repre-
"eating the dead weight of the reciprocating parts.

Correction for Inertia.


It is convenient to make one more scale, to show pressure per
square inch of piston. The piston area being 491 square ins.,
divide the total pressure reading by 491 to obtain reading per
sq. in.; stepped off at s/.

The indicator card for the top of the piston is set out by the
unit pressure scale at s/, and appears as EQXHB, the bottom of
diagram touching the base AB. Similarly FPGA is the card from
the bottom of the piston. Now, while QXHB is being drawn by
the indicator on top side of piston, A F R would be formed by
that connected with the bottom side, and the -effective pressure
will be the difference of these curve ordinates. Deduct those
at F from those at H, and the result is the curve WR. So also
VN is the curve of effective pressure on the bottom side of the

Now the actual total pressure to be carried forward to the
crank pin will be, during the first half of the stroke, less than that
on the indicator diagram by the amount required to set in motion
the reciprocating masses, viz., their inertia; and during the second
half of stroke the indicated pressure will be increased by the
backward pull needed to absorb inertia. Briefly, then, the c top'
card loses by the area ANS, and gains by SBP, the resulting
pressure area being NXWP; and similarly the resulting area for
the 'bottom' card will be P/VN. Setting up the resulting
ordinates on the straight base AB, we have the curve A£*/B for
the top and B <?/A for the bottom of piston, the total pressures
being written on each ordinate; and in order to equalise the
areas the cut-off in top diagram has been placed at '3 and in
bottom at '6 of stroke, the dead weights having to be supported
in the latter case.

We must next distinguish between reciprocating and rotating
parts, for only the former cause inertia force. The piston, piston
rod, crosshead, and smaller end of connecting rod are recipro-
cating weights, but the larger end of connecting rod is a rotating
weight As regards the connecting rod itself, about two-thirds
may be called reciprocating and the remaining third reckoned as
a rotative weight. The reciprocating weights directly affect the
indicator diagram, and the latter must be altered, by increased
compression or later cut-off, until a fairly even pressure is

676                      Curves of Crank Effort.

obtained.    The revolving parts must be balanced by opposing
weights on the crank shaft.

Curves of Crank Effort.—If the crank be on either dead
centre, there is no tangential or turning effect produced by the
steam pressure on the piston, all such pressure being received
upon the bearings. When the crank is midway between dead
points the whole piston pressure is transmitted tangentially, and
there is no pressure on the bearings except that due to dead
weight. Between the two conditions part of the pressure is
transmitted tangentially and part normally. But (p. 491) the
polar curve proportionally represents tangential crank pressures,
other things being equal. Divide then jo, Fig. 660, into tenths
and measure the radii vector of the curve UK in terms of these
divisions: the numbers obtained will represent the virtual crank
arms in relation to pressures transmitted along ABO. Taking the
total pressures from A to B, multiply each pressure ordinate by its
virtual crank arm, and the result will be the tangential crank
pressure for that position. Setting out these results radially, with
the crank circle JKLM as a base line, we obtain the two curves of
crank effort ighji. and L-klm] for the top and bottom of piston
respectively. These again are better understood on a straight
base, so the base JK is stepped out at co, KL at OD, and the
radial ordinates transferred as vertical ordinates on the new base
CD. Curves CHD and D/C are thus arrived at.

Combination of Crank Effort Diagrams.—Though
the fly wheel may equalise very tolerably the crank effort, there is
still the difficulty of starting when the crank is at either dead
centre. This is not a material difficulty for a factory engine
which has only to be started twice a day; but in locomotive and
marine practice it would be a very serious obstacle. In loco-
motives two cranks at right angles are employed, as at i a, Fig.
447, p. 486, while in marine practice it is usual to place three
cranks at 120° mutually. The latter gives the best conditions, but
the advantage of both will be made clear in Fig. 661.

At (a) two cranks set at 90° are each supposed to have effort
curves, as in Fig, 660. Plot these with relation to the respective
cranks, A A being the top curves, and BB the bottom ones. Then
the curve of total effort may be found by super-position, that is, at

SCALfi.   OP    TOTAL

/«    ACCfcU. «CAU£ X §£££



TOTAL   DEF/C? =   3jSl^9 + '4-956O »   687OO rr.Ui


Weight of Fly Wheel.                        679

every radius the ordinates of both curves are added to form
the resulting curve cc. An average circle is struck, and shewn
dotted; and a clear conception of the more even turning move-
ment is then obtained.

Three cranks are set out at (#), Fig. 66j, and the like process
followed. The same letters are adopted throughout, and a more
regular turning movement results. The differences between the
c c curve and the dotted circle may seem little better than before,
but they form a much smaller percentage of the effort ordinate.

Calculation of Fly-wheel Weight required.—The
crank radius, Fig. 660, being i| feet, the circumference of the
crank circle is exactly n feet. In Fig. 662, let ad be 22 feet,
and let it be divided so that ae and hd are each 2| feet, and <?/
/£, and gA, are each 5 J- feet. On ef and g h set up ordinates
of crank effort on the up stroke, and on fg of that on the down
stroke, those on a e and • h d each representing half the down
stroke effort. Now take the mean of the ordinates on ef:
dividing the base into 10 parts, measuring at centre of each part,
adding the ordinates and dividing by 10 : the result is 29,500 Ibs.
The mean of the ordinates onfg is found similarly to be 25,000
Ibs. Adding and dividing by 2, gives 27,250, the mean effort
for the continuous diagram a d. Draw / k at this pressure
above ad.

Now the areas, A, c, &c., shew surplus work, while the crank
travels from I to #z, and from n to p respectively, while the areas
B, D, &c,, shew a work deficit between m n and p q. The fly
wheel must absorb the work A or c, and give it out again at B or D,
and thus tend to equalise the crank effort. The mean pressures
and distances traversed have been measured at A, B, c, and D, and
are shewn by work rectangles. The total surplus and the total
deficit of work per revolution ^ are each found to be 88,700 foot
pounds, and the greatest of the four work areas A, B, c, and D, is
D, or 49,560 foot pounds. This is the amount of energy
which the fly wheel must be able to deliver, such delivery
decreasing its velocity, while the absorption of energy will in like
manner increase it But the heavier the fly wheel, the less will be
the fluctuation of velocity; and the problem is to find the weight
of wheel which will absorb the surplus energy and re-deliver it

68o                        Weight of Fly Wheel.

keeping the fluctuation of velocity within a certain percentage of
the mean. Let v — mean velocity, and let

Total fluctuation of velocity = -7 of v


then the value of k 'depends on the regularity required, and
may vary from 100 for very steady driving, to 20 where constant
speed is of little value. With feet and seconds units, let z^ be
maximum velocity and z>2 minimum velocity of the fly wheel at its
mean radius, consequent on absorbing and delivering the given
^energy, and let E represent the energy area, or the 49,560 foot
pounds of Fig. 662, while the velocity falls from v^ to v%.

j      z*> (V ~ z>22)
Energy delivered = — ^ - ^•


where w is the weight of the fly wheel. But this energy is equal
to the area E,


NOW Vl - 7/2 = V   X  -                   flj + V2 = 2 V

27T RN

and     v = — - -

R being radius of gyration of fly wheel.
Putting also the fly-wheel weight in tons,

-        (#! + V%) (7^ - V^) 2240         2 V2 X 2 240

&32'2 ^ 60 x 60               E^


Generally R may be taken at centre of * rim section, but if
great accuracy be required, assume a fly-wheel section, replac-
ing the arms by a thin disc of equal weight : then, moment of
inertia ,of volume (second moment) of a solid of revolution, or
JT= volume xR2, which is the third moment of the generating

Steam Port Area.


area (A),  or of half the fly-wheel section.     Bat volume = first
moment of A

3rc^ nioment of A

•D2 _    v

"~ vol.

ist moment of A

the moments being found graphically at p. 845, If inches have
been adopted, R must be changed to feet when inserting in the fly-
wheel formula. Also #must be measured at radius R. If W does
not now agree with the calculated weight, the section must be
altered, and a new calculation made.

Area of Steam Port. — Practice has decided certain average
speeds of piston in particular cases, and the following list has
been thus deduced : —

Locomotives .........................................  1000

Marine engines .......................................    700

Horizontal engines ....................................    400

Pumping engines   ...................................    130

To attain this speed, we must not endeavour to pass the steam
through the steam port at a greater speed (according to Rankine)
than 100 feet per second.

Ratio of

cylinder area


port area         speed of piston in ft. per sec.

If the port be much contracted, a lower piston speed will be
attained than that intended. We shall close this chapter with
some practical examples, together with a few remaining points
of theory thereupon raised.

Horizontal Compound Pumping Engines. — Figs.
663-4-5 are views of a pair of compound engines designed and
built by the East Ferry Road Engineering Works Company,
Millwall, serving as examples of the Stationary or Land engine.
The engines are used at the Millwall Docks for pumping water to
hydraulic accumulators, under a pressure of 750 Ibs. per sq. in.
The bed plate H is in two parts, and supports the high-pressure
cylinders BB, the low-pressure cylinders A A, the pumg cylinders
zz, and the crank shaft bearings mm. The H.P. cylinder is
supplied with a main valve D, and an expansion valve E, of Meyer
form, the valve spindles being lettered respectively G and F. The


Compound Pumping Engines.


valve c for the L. P. cylinder is double-ported, but not relieved
from steam pressure at the back. The piston rod YY, being pro-
longed, forms the pump plunger, its sectional area being half that
of the pump piston, for reasons to be explained in the next
chapter. J j is the steam supply pipe to the H. P. cylinder, K its
exhaust, as well as supply for L. P. cylinder, and the L. P. cylinder
exhausts directly into the condenser, as will be seen in Fig, 665.

The surface condenser NN consists of a.rectangular casing, con-
taining a nest of tubes. Cold water being allowed to flow in
through the pipe M, passes through these small tubes, first through
the top half, returning through the lower half, then out by the
pipe L. The exhaust stearn distributing itself outside the tubes,
becomes condensed, is taken away as water by the air pump Q',
and delivered to the hot well u; then by the feed pump v to the
boiler. The air pump bucket has four valves at T, fixed foot

Three-cylinder Marine Engines.


valves s, and delivery valves R to prevent the water returning.
The bucket is actuated by the bell-crank lever ww, connected by
links to the crosshead x. The connecting rod e has a long fork
to clear the pump barrel; it is also light in construction, its sole
duty being to transmit equalising energy to or from the fly wheel,
in addition to the power required to work the valves. The pump
suction pipe is at d d, and the delivery pipe at bb, but full ex-
planation will be left to the next chapter. We must not omit
to mention the hydraulic governor g, the invention of Mr.
C. R. Parkes, M.I.C.E., which has given great satisfaction in
its working. The flying balls are driven from the engine in
the usual manner, but the sleeve opens a small D valve to
hydraulic pressure or exhaust, according to whether it rises or falls.
Nothing takes place until the governor has attained a speed of
15 revolutions per minute, when high-pressure water is admitted
into the cylinder /£, and the ram / is pushed downward, thus also
pulling down the strap k and raising the weight /. The conse-
quence is that the pulley f, on the expansion valve spindle, is
rotated so as to increase the lap of the Meyer valve and secure an
earlier cut-off, and the action will continue until the speed of the
•engine has returned to the normal, when the governor sleeve will
fall, open the D valve to exhaust, and allow the weight / to lift
ram j to its original position,

Triple Expansion Marine Engines.—Figs. 666 and
£67, Plate XVII., are two views of the triple-expansion engines of
the Pacific steamer Iberia, designed and constructed by Messrs.
David Rollo and Sons, of Liverpool. The bed plate y, in three
pieces, carries the left-hand standards; the right-hand standards K,
KJ, and Kn, being built upon the condenser v. Cylinders.—The
H. P. cylinder A is 33 ins. diameter, B the intermediate is 54 ins.,
" and c the L. P. cylinder is 88 ins.; while G, H, and j are the
respective pistons, of conical form to combine lightness with
strength, and each having a stroke of 60 ins. To minimise the
number of spare parts, the cranks YVY, connecting rods zzz, piston
rods DBF, eccentric s and rods STU, links r, gudgeons zz, crossheads #, •
and pump levers jk, are all made respectively interchangeable;
only a small alteration occurring with the rod D, which must have
the tail or upper part cut off. Valves.—A piston valve b is



adopted for the H. P. cylinder, packed with flat rings, but dis-
tributing steam like a D valve. To save space two piston valves c
are supplied to the I. P. cylinder, as seen by the valve rods 3 and
4, connected at their lower ends by the strong crosshead 5; and
the L. P. valve d is double-ported, while being relieved on its
back by the hollow piston h. Piston £", with steam pressure
underneath, supports the weight of valve //, and the L P. valves
are similarly supported by pistons within x and w. Relief valves
UU) on the cylinder covers, are wing valves weighted with springs,
serving as safety outlets for condensation water, which might
otherwise break the covers when the pistons moved. The H. P.
and I. P. slide valves, being vertically above the crank shaft, their
eccentrics are set to lead the cranks, but the L. P. valve is moved
by the rocking lever Q R, and its eccentrics must therefore follow
their crank (see p. 646). The radius link is formed of two
plates, having the die between and the eccentric rods outside,
thus enabling the pin centres to be coincident when in full
gear. The steam reversing cylinder / has its rod u coupled to
the weigh bar lever spy which, through the drag link qr, moves
the link r to fore or aft positions; expansive adjustment is given
by the screw q, and 9 is the valve lever for cylinder /. The
exhaust steam passes to condenser v through standard K ; and air
pump x and circulating pump w are worked from crosshead v by
levers jlk. 7 and 8 are oil pumps, and 6 an oil reservoir with
gauge. The cylinders are jacketed at sides, top, and bottom;
and drains connect to tanks which shew the water used.

Condensers.—The advantages of condensation having been
discussed theoretically, we will now describe the three principal
methods of realising those advantages practically.

The Jet Condenser, Fig. 668, applied to most land engines,
consists of the condenser A, where exhaust steam E is met by a
constant spray of cold water from injection cock G; the air pump
B, worked from the engine piston rod; and the hot well c, from
which the condensed steam and water is taken to feed the boiler.
In order to make B'S action continuous, there, is a suction valve s
and a delivery valve D at each end of the cylinder.

The Surface Condenser; Fig. 669, avoids the mixing of cooling
water with the stearn directly. Formerly, if such water were dirty

" r' k tt

- h } If



---.....—>-».--. JRifl


I       J       *   '     *

»          ft      K       ,   "   # '

•, •    i -   »     ; ''f -trv   * »fc»

Compound Locomotive.


horseshoe bearings have water passing in'and out at j j. K K are
lifting eyes.

The Stern Tube, Fig. 673. A is the tail shaft, tapered to fit
the propeller, where it is keyed and gripped by a nut and split
cotter. A renewable muntz-metal sheathing D is rolled on the
shaft, and gives a smooth, non-corrosive working surface. The
tube B, bolted to the water-tight bulkheads at H H, and supported
by the stern frame at c, has a bush E in which are placed staves
of hard wood (Hgmim vita), being the best bearing where water
is the lubricant. At the other end a stuffing-box, formed by the
neck ring F and gland G, prevents water entering the tube.

Compound Locomotive.—The general arrangement of a
locomotive being well known, one good typical example will here
suffice. The example chosen serves to illustrate the ordinary
'inside cylinder7 engine, having cylinders within the frame, the
only main difference being the arrangement of steam pipes. It
also shews one of the most successful adaptations of the com-
pound principle to locomotives.

Figs. 675-6-7, Plate XVIII, are views of a Compound
Express Locomotive for the North-Eastern Railway, on the
'Worsdell and Von Borrie' principle. The main frame consists
of two plates LL, a cross stay Lt, and buffer beams MM: the front
beam carrying the buffers N N, draw hook b, and coupling screw c,
while the back beam faces that of the tender Q. Between M
and Q are placed buffers p, pivot 50, and safety links 88, the pull
being taken by the draw-bar 6. 35 is the foot-plate, 19 the cab,
to shield from the weather, 34 the platform, and y the splasher for
the driving wheel: //"are lamp brackets, and dd lifeguards. The
cylinders A and B are bolted between the frame plates, and slide
calves aal are placed above the cylinders to suit Joy gear, whose
various links z, Y, x, and w are explained at Fig. 640. There are
four slide bars qq to each cylinder, and two motion blocks rr\
n and/ are the piston rods, and mm the connecting rods. The
wdgh-bar shaft s is moved by a hand-wheel and screw at 0,
coupled to lever / by the rod u. E E are the driving wheels, and
F F the trailing wheels, with j and K the respective axles : the
former is known as the crank axle, and in the N.E.R. example is
turned throughout The wheel centres are of cast steel, but the

690                       Compound Locomotive.

tyres are rolled weldless and fit into annular grooves in the wheel
rim, to resist centrifugal force. The front end of the frame is
supported by a trolley or bogie, which permits certain side move-
ment when travelling round curves. H H is the bogie frame, with
stay rods TT, cc the bogie wheels, and GG the axles. A block or
die 43, curved to a radius from 42, is held by the pin D ; and
guides 44, similarly curved and forming part of the bogie frame,
ride upon the die. If 43 were rigid, the bogie would only swivel
round 42, and would only adjust itself to certain curves; but the
freedom of 43 on D permits a further angular movement, and
the virtual centre 42 is therefore variable. The buffers u u limit
the lateral deviation, and springs gg return the bogie frame to
central position.

Laminated springs> h and 13, and helical springs, 45, placed
between frames and wheels, lessen the shock due to inequality of
permanent way, and the necessary vertical sliding is met by
providing special bearings, S, x, and 14, termed axle boxes, for
the bogie, driving, and trailing axles respectively. R, w, and 15
are the guides in which the boxes rise, and z is a wedge to take up
wear in the main box.

A hand brake r i is used in emergency, but the regular work
is performed by the Westinghouse compressed air brake. The
steam pump 51 fills the main air receiver 5, from which auxiliary
receivers—one to each carnage, and one, 46, for the engine—are
further supplied. From 46 pipes are led to the cylinders 4 4,
and the air pressure moving levers 2 2 put the blocks 3 3 on the
tyres. Upon exhausting, 3 3 are released by springs. When
ascending steep gradients, sand is driven between wheel E and
the rail by means of a steam jet from pipe /, the sand passing
from sandboxes jj, down the pipe k. Cylinder cocks 18 act as
relief valves, and are opened after the engine has stood some

Boiler 20, and firebox 21. Very little description is needed
beyond that at p. 335. Girder stays 53 are of cast steel, and 52
are long stay bolts. A firebrick arch 23 deflects the current of
heated gases over the box, and the ash pan 2 has doors or
dampers, both before and behind, for regulating the draught.
The draught is 'induced' by the exhaust steam escaping at the


Compound Locomotive.

blast pipe 39, gradually contracted towards the orifice to cause the
necessary velocity ; and the smoke box 22 must be air tight, so its
door 33 is provided with two handles, one for turning the tongue
catch, and the other for tightening the screw. A jet of steam
from the blower 40 causes draught when the engine is standing.
The steam regulator, 56, has two slide valves worked from, handle
34, the main valve 27 being treble-ported, and the 'easing' valve
28 double-ported and small. A pin 55 connects both valves to
the gear; but the hole in 27 is slotted, so that when opening, 28
is first moved (easily, being small) and a film of steam admitted
'between the main valve and its seat. Next, 27 is caught by the
pin, and, on account of the relief just given, can be moved
without difficulty.

Under ordinary conditions steam first enters the H. P. cylinder
B by the pipe 27, exhausts thence to the L. P. cylinder through
30 and 28 (the whole pipe forming a receiver of a capacity equal
to B), and finally leaves by the blast pipe 39. But if H. P. crank
be on a dead centre at starting, steam must first be admitted to
the L. P. cylinder A, and yet be prevented from entering B for
fear of blocking the piston. Outside the smoke box a valve box
61 is fixed, having a starting valve 59 opened by a rod from the
foot plate when required, but at other times kept closed by a
strong spring. Pipe 41 takes steam from the boiler to 61, and 57
carries it away to the main pipe 28, entering at 29; and a piston
62, fitting in the valve box 61, is connected to the rod 60 for the
purpose of lifting the flap or intercepting valve 58, which is
normally open. When the driver wishes to start, he opens
regulator valve 27, and if the H. P. piston refuses to move, he
pulls the small lever which opens 59; and steam, wire-drawn to
half pressure, enters 61, moves 62 clear to the left and closes 58,
then passes by 57 and 29 to move the L. P. piston. Once the
engine moves, steam enters the H.P. cylinder by 27, the proper
path, exhausts by 30 and 29, and acting on the large area of the
flap 58, opens it, and once more valve 59 is closed by its spring.

Instead of feed pumps, injectors are now favoured for feeding
locomotive boilers, and two of these, 12, 12, are supplied. They
draw from the tender through a strong rubber,pipe, and deliver
through the clack box 25, in which is a non-return valve. A




Tractive Force.


double spring-loaded safety valve, 31, is placed over the firebox.
The valves are inverted cones, fitting easily, and either centre-
point can be lifted by the lever 32 to test the working. A safety
link placed within the spring holds the lever in case of breakage.
71 is the steam whistle; 70, a lamp bracket; and 72 the chimney,
of cast iron. 38 are lubricators for the steam chests.

Tractive Force of a Locomotive is usually taken as
the mean pull exerted on the moving train, and may be estimated
from the principle of work. Thus :

Work given by Steam = Work done on Train.

Total mean pressure )
in both cylinders    J

x stroke = Tract, force x

/ half wheel
1 circumference

2 x "

/ = T x TT r

.-.   T =

The tractive force for any particular starting position can
only be found by first ascertaining the crank effort, for that
position (E) ; then, by moments :

2 r

Of course, the greatest value of T must not overcome the
adhesive force, or slipping will occur (seep. 571). The tractive
force required is given at p. 569.

Boiler Fittings.™- Boilers having been described at pp. 330
to 339, it remains to consider the principal mountings with which
they are fitted, (See Appendix I^p. 755; Appendix //.,/. 833
and pp. 899-905; Appendix III., p. 918.) (See also p. 1148.)

Safety Valves.— Lever-loaded valves, p. 482, are not now
in favour, on account of the fear of explosion due to sticking.
Directly-loaded valves may be either spring or weight loaded.
The former has been shewn at 31, Fig. 675, Plate XVIII ; and a
dead-weight valve is given at Fig. 678, as applied to stationary
boilers. A casing A, containing the weights, is hung on a cup-
shaped valve resting on the conical end of the pipe B. The
figure shews also a low-water float c and a high-water float D,
which raise rod E whenever the water falls too low or rises too
high respectively. Marine valves are spring-loaded, and the

Boiler Details.

Board of Trade Rule gives half a sq. in. valve area tor every
sq. ft. of grate surface. (See p. 902.)

Mudhole Cover.—Manhole covers are merely flat plates
covering the raised mountings shewn at Figs. 310 and 311 : mud-
hole covers. Fig. 680, are more perfect mechanically, the oval
plate A being kept closed by the steam pressure, and further
secured, by bolts, to bridge pieces B B. The oval shape permits
the plate to be entered narrow-ways, after which it is adjusted
into position.

Pressure and Vacuum Gauges.—The Bourdon Gauge,
Fig. 681, is now generally adopted for both purposes. Within
the casing A is a curved -tube c, of flattened section, as at D : it is
open to pressure at B, but blind at E. If the pressure increase
above the atmosphere, the tube distends, and point E moves out-
ward ; but a decrease of pressure below atmosphere still further
flattens the tube, and point E moves inward Both movements
are transmitted to sector F, which turns, by a pinion, the pointer,
thus multiplying the motion; and a hairspring on the pointer axis
takes up backlash. The graduations are made by experimental
comparison with a mercury gauge.

Injectors.—There are two ways of feeding a boiler with
water when under steam : (i) by a pump either driven from the
•engine, or steam-driven and self-contained, then known as a
f donkey-purnp j* (2) an injector may be used. Pumps will be
treated in the next chapter. The injector forces water into the
boiler without the intervention of moving mechanism, and the
only loss is that due to fluid friction, while the delivery of hot
instead of cold water is a gain, not to speak of the diminished
strain on the boiler. In Fig. 682, A is a section of the instrument,
and B shews its application. The injector represented, being
non-lifting, must be placed at the tank bottom, but its parts are
essentially the same as those of other injectors, c is the water-
cock, D the steam cone, E the combining cone, F the overflow
pipe, and G the delivery pipe. H is the steam cock, and j a
non-return clack box. Cocks H and c being opened, water is
forced into the boiler by the steam; and it long remained a
surprise to engineers that steam could feed water against its own
pressure. The explanation is this: the velocity of efflux of steam



is some 16 or 18 times that of water at the same pressure, and
the jet of steam escaping from D is so suddenly cooled by the
tank water through c, that it has not time to reduce its velocity
to that due to it as water, and therefore succeeds in piercing the
boiler water, carrying the tank water with it. An overflow takes
place at F when first starting, which, however, ceases when cocks
c and H have been mutually adjusted.

A lifting injector must permit of regulation at the orifice D
and ring orifice E, for the conditions of vacuum-forming and
water-forcing are quite different, and the former must be first
satisfied, after which the latter may be met without disruption
of the water column, An ingenious method of automatic lifting
injector is now in operation, where the throat E is split longitu-
dinally, and one half hinged near the annulus, the' flap nozzle' thus
formed also causing are-starting, should the fluid tend to disunite,

Slightly altering the cone proportions, and giving the water a
few feet of head, produces an injector workable by exhaust steam ;
but, on account of the variation in pressure, the flap nozzle must
be provided.

Other Mountings for the boiler are: a blow-off cock near the
firebox bottom, gauge cocks about 3 ins. above and below the
water line, fire bars and bearers, furnace doors, rnud plugs, fusible
plug in furnace crown to melt in case of overheating, and thus
•cause the fire to be extinguished, clack box (j, Fig. 682), damper
for regulating draught, a filling branch when no other hole is
-convenient, and sometimes a scum cock. (SeeApp. •//.,/. 902.)

Combustion.—Combustion or burning is rapid chemical
combination accompanied by heat and sometimes light. If con-
siderable noise be caused it is termed an explosion. During
•combination, heat is produced equal to that required to separate
the same elements. The separation of carbon and hydrogen,
and their re-combination with oxygen, is what the engineer
needs to understand, so \ve will consider the burning of a simple
hydrocarbon like marsh gas, shewn by the formula :

Marsh gas 4- Oxygen. = Carbon dioxide -t- Water (steam)



a case of complete combustion, for no single element remains.



Taking the atomic weights of C, H, and O, as 12,  i, and 16
respectively, we have :

Marsh gas  +   Oxygen   = Carbon dioxide +     Water

(12 + 4)   +2(16x2)==   {i2 + (i6x2)}   +2(2 + 16)

that is,      i61bs.     +    64lbs.    =         44lbs.         +    36 Ibs.

or,             ilb.      +      4 Ibs. gives    275 Ibs.         + 2-25 Ibs.

Again, i Ib. of carbon burnt to C02 gives 14,500 thermal
units, and i Ib. of hydrogen burnt to H2 O gives 62,032 units,
In i Ib. of marsh gas there is £• Ib. of carbon and \ Ib. of


•. -fib. Carbon + O      gives  14,500 x £ = 10,875
Jib. Hydrogen + O  gives 62,032 x \ = 15,508



Experimentally we obtain a total of  23,582 units, or   2801
units has been required for decomposing the C and H.

Good dry bituminous coal contains on the average, by weight,

Carbon, 83-5 %      Hydrogen, 4-6 %      Oxygen, 3*15 %

the remaining 8*75 % being Nitrogen and Sulphur, inactive
elements. Taking 100 Ibs. of fuel the 3*15 Ibs. of oxygen is
already united to -J- x 3*15 = '4lb. of hydrogen as water, and the
hydrogen does not assist combustion; so we have left:

83*5 Ibs. of Carbon

4*2 Ibs. of Hydrogen

Now 12 Ibs. of C unite with 32 Ibs. of O, or as i : 2'66 ; and
2 Ibs. of H require 16 Ibs. of 0, or as i : 8.

Ibs. of O.

.*. 83*5 Ibs. C require 83-5 x 2*66 = 222
and 4*2 Ibs. H require   4*2 x 8      =   33*6

Total weight Oxygen for loolbs. coal = 255'6 Ibs.

or 2-5 Ibs. of Oxygen is needed to burn i Ib. of coal.    But air is
composed of 77 parts Nitrogen to 23 of Oxygen, by weight.

/.   23 : 100 :: 2*5 = 10 Ibs. of air per Ib. of fuel.


698          "                  Forced Draught.

Again, we have per Ib. of such fuel '835 Ib. of C and '042 Ib.
of H,

Heat units.

.-. -835 Ib. Carbon + 0      gives  14,500 x -835 = 12,107
•042 Ib. Hydrogen 4- *O gives 62,032 x '042 =   2,605

Total units            .    14,712

By careful laboratory experiment one Ib. of such coal is found
to have a calorific value of 14,701 thermal units, and evaporate
15 Ibs. of water at 212°. Also 12 Ibs. of air are required per Ib.
of fuel. (See pp. 906 and 1148.)

In actual practice considerably less beat is developed, and the
evaporation is good at 10 Ibs. of water, being commonly 6 or 8.
Also 24 Ibs. or 312 cub. ft of air are required, with natural
draught^ to dilute the gases, and allow the air to reach the fuel.

Forced • Draught.—The essential advantage of forced
draught lies in the fact that a smaller dilution of the gases can
be allowed, i81bs. of air per Ib. of fuel, or only i|- times what
the chemist requires. In consequence, a higher temperature is
obtained, the grate and heating surface being much more efficient;
and thus a smaller boiler will serve the purpose, a great advantage
in torpedo boats.

The air must not be solely fed through the fire bars, or a
tongue of flame would meet the stoker whenever he opened the
fire door. The closed stokehold, the earlier method of solution,
places the stoker in a plenum of air at a moderate pressure, which
enters the furnace as usual. The later method, the closed ash-
pit, requires a box-shaped fire door, into which air is fed as
well as to the ashpit; but the air to the latter is at a much lower
pressure. The air from the box door passes to the coals through
holes in the baffle-plate, and the supply is cut off automatically
whenever the door is opened. Both methods still have their
advocates. The pressure is caused by a fan. (See App* 12.,p. 907.)

Waste of Fuel is largely due to formation of smoke and
incomplete combustion, the carbon partly being burnt to CO.
Alternate or continuous firing, by careful men or mechanical
stokers, and a sufficient supply of air, are the only remedies.
The gases also pass up the chimney at a greater heat than 600°,

-   %j   9;   ^    %,         jt   *   %

*   *      .

£   f   -


Methods of Ignition.

But one detail has caused some trouble to all inventors, the
question of igniting the explosive mixture without escape of gas.
Three methods have been used:—(r) Plame ignition, where a
portion of burning gas is carried through an aperture in the slide
when the latter is just closing. This method has been used
extensively, but occasions frequent misfires when the small aper-
ture becomes carbon coated. (2) Tube ignition, Fig. 682 a.
Here the blind tube A is kept at a white heat by the bunsen
flame c, supplied with gas from B, and whenever the timing valve

E is opened by the spring G, the charge, which has been com
pressed into the ignition chamber D, then ignites. F is the boss
of a lever which keeps valve E on its upper seat, and allows the
contents of the tube to be cleared through hole T. Small engines
have no timing valve, ignition only occurring when the charge is
compressed into the tube. Iron tubes have to be replaced every
fortnight at the latest (3) Electric ignition was adopted in the
Lenoir engine, but in a faulty manner. The current from battery
L was intensified by the coil K. It passed through insulators at
M M, and by platinum points through the cylinder N, the circuit
being closed by the crosshead j, causing sparks at MM, The
covering of the platinum points with carbon or watery vapour

The Simplex Gas Engine.


•was the cause of failure.* In the Simplex engine a constant
•srrower of sparks takes place in the chamber x, the current
passing through the insulator u and back by v. In the figure the
cylinder is being charged from s, through Q, but when the slide
moves to the right, R connects w with x, and ignition occurs
•with certainty. (See App. /,/. 773, and App. IV., p. 963.)

We may now describe the SIMPLEX ENGINE (Systeme Delamare-
Deboutteville et Malandin), Figs. 683 to 688, as a type of a
•well-designed gas engine. A is the cylinder, supported on the
•bed plate H, and surrounded by a water jacket B, which also
protects the slide casing and exhaust outlet; N is the mixing
•chamber, and c the piston or plunger. D is the connecting rod,
E the crank, F the balance weights, R the crank shaft, and G G the
'fly wheels, having a pulley p attached for driving purposes. Pipe j
:is always open to air, and the gaspipe K admits gas when cock L
is opened. But such gas is only allowed to enter the cylinder
.at proper times, viz., when the charging valve M is opened by
projection h on the slide spindle g. As the cycle occupies two
revolutions, the shaft Q (which moves the slide d backward and
forward through the disc crank/) makes two rotations to one on
the main shaft, and the wheels at R and s together have a velocity
Tatio of 2:1. The charging and ignition having been described
*the governing and exhaust arrangements remain. Taking the
former, shewn in Figs. 687 and 688, the method adopted, as in
-other gas engines, is to cut out one or more chargings when the
•engine speed increases. Upon the spindle h is a small tapered
6 rocker '/, &&& when this is allowed to catch the stem k of the
•charging valve the latter is opened The governor is a pendulum
n /, whose lower end is lifted to the right by the rocker /, and,
being allowed to return freely, its time of fall is invariable.
Noting that the rocker/ is constantly depressed, as in Fig. 688,
•by a spring, suppose engine speed to be normal, and / to be
•moved to the right, lifting the pendulum. Returning, the pen-
dulum bears slightly upon the rocker, catch m lifts/ to the hori-
zontal, and the valve is opened. But if the slide travel too
quickly, m misses/when returning, and the result is a 'misfire,'

* M. Deboutteville. iodines to the former, Prof. Wm. Robinson to the
latter (muse.

Y    MM.  MATTER cr    cii.,  ROUEM.

Petroleum Engines.


as shewn in Fig. 688. The pendulum may be adjusted to the
greatest nicety by raising or lowering ball n. The method or
opening the exhaust valve is seen in Fig. 686. A cam e on the
shaft Q lifts the lever T, pivoted at u, and, through rod v, the
'crocodile jaw' w; thus raising the valve against the springs a a.
w has a shifting fulcrum at x, giving a larger leverage at first, and
a quicker opening afterward.

Fig. 684 shews the indicator diagram obtained, which still
further illustrates the Otto cycle. One difference in the Simplex
working is noticeable; the mixture is over-compressed, that is,
a small return motion is made, after leaving the dead point,
before ignition occurs, and the force of the explosion only reaches
the crank when it is in a better position, viz., at 15° from dead

For the best economy, gas engines should work with 'poor
gas,' as produced by the Dowson plant in England, and the
Buire-Lencauchez in France: the latter is used in conjunction
with the Simplex Engine. Rich lighting gas is expensive for
large engines. (Seefjb. 911 and n$i.)

Petroleum or Oil Engines, like gas engines, are of the
internal combustion type. Petroleum occurs naturally in Russia
and America, but is also obtained as paraffin by shale distillation.
It is highly complex, consisting of several liquid hydrocarbons
having different boiling points: thus, when heated, giving off
first the lighter oils, then the burning and lubricating oils, and
lastly paraffin wax or vaseline, leaving a residuum. The light
oils, including benzoline and naphtha, are dangerous, flashing at or
below 73° F.; while the heavy or lighting oils, like kerosene, are
thoroughly safe, resisting the flame of a match, or even the
electric spark. But the heavy oils are difficult to prepare for
the motor, where they are to be intimately mixed with air to
form the charge: if vaporised at low temperature, a troublesome
residue is formed, while gasification at high temperature produces
also tar.

In 1838 Messrs. Priestman Bros, acquired the Eteve patents
(where spraying with air and evaporating in a hot chamber was
first proposed), and after considerable experiment produced the
first practically successful engine working with safe oil, doing for


f^f»     Jj^i     I             M»;**i

-«it  ,«***Jf**4   f   *^    t«fr>

^'-i'/f'/    V"/f   ''   '*    S^^.f  f


-f • *-  ** v'| "•

" '*   * t f ? / "    '

j     *4^      t'     i,      t   ^    '

»> *     * *?     , A ^

,     ,„     v      ,,         .   ,     t
, r>         *'       i    •         ' t           >i

i   '   ' /     • "   f

•    *                  ;    ,|

V,       ' ""   .   i          V   i'         •,'?,«>      f I

.   .   J*     /*      »   ."        '       r   *  f                 '      ."     %,*/*/

f    ]   ',         f ' i       '  '. f,
I   #,>.-,'*;      or-.

. J         ' ' *               5 !

« <   1;   ,    * 1  1  t>    ,     f     'IV,,'       '•       'V*  »    /     J   7    ^*t'    *  '     lu  *,   W, •*.*//     **'*?!

J|f»j  t/Jj»    ^f   « '«  i »/-     /     ^'fKj.J/1    n1    |,JJW5?    I//    <l^                             **-*?    *

*H M1* ^   *   M #^t '    '*^    Af'i^ t *:*<* «^   ^ <*    I/ A^ff,   ##*   4 l^irf»*|   ^2<*#   fc^# «'fi f $

0*.»f     |r*«|t ^*t#t«l    f^    *l^*>l        1,lfl|ff||J     ll^f^'flf      ^t-    *      «4    >!///   1**  -^S     'f «        *'»      J     v   /    «     «

flilf^  fcf^rt   ffrffl^   t*"J*!Hr   tf^fci,    ^Vh   ^''5<        '^      '* ' ?>      •        x     i

fl                        l!i«i»ffr    -»*|Si     v    ^,4^   | '   i    «j        '  ,      «       '.              ;'

m/it ilr  *  i  t.t ]f    s'.'t it ,?'-/•,
«lr                 * oOr> t

4*ll|    t f  ^A*t<^/   4?   : nt  |   , '    '   *    '^            '          *         '   '     4  '     ** l   'r

fl'   1^^*'",    •/}(<.**»         «    **•     |   ^*r»    , 1     f      ,        -     «       A*    y*

1 *•

f f 4 I   f    \ 1^*1  *rt «t * r/'i  J  .   r* "A      , ; ** *    i.i'-1'   *%<    : <^>> »   t   .        * fl '< ft

4^''^**'*"                                                  *|                    '                                  <

tllf           i^vf  f.#i   I-    , t  $ ^  ,w *   i • *       |j^^       »•% t   '^^ 4 f* '.t   i4

()^t^r, *!•,*!», ^$i «   M*»H|,,  * !.^^ f|4*4i'4l?"      ,^f * ^X*   v^ ,/-< •<>*«?*

i|»  j> «%                      Ii4               i.v  ^^ IA|.||   4 ; t*^  V* I   -.'s*^

IM                             f^i            ^ «t^  in |-i*,t'/i*iu«   t »»^i*ifi

^f/   II       t*   Jfef             f(««ir   ^«   fttf*«,*J


Other Oil Engines.                          709

is placed above j and is connected to it by two pipes (for oil and
air). When the pressure in the two tanks becomes equal the oil
runs into j by gravitation. Lubrication is effected in the usual
manner, at all parts of the engine except the cylinder j the oil
condensed within which is ample for the purpose.

Several forms of oil engines are now made by other firms, but
none spray the oil. In some, liquid oil is evaporated in a hot
chamber, forming vapour and gas, which is mixed with air and
fired as usual; and it is said no deposit occurs in ordinary
working. In others, perfect oil-gas is produced, and then ex-
ploded with air, but the engine must be often cleaned from tarry
matter. (See Appendix II., p. 915 ; also p. 1165.)

Oil Engines for Motor Cars.—The requirements of
motor cars have developed an engine of very small weight
using a light benzine oil called 'petrol.' Further information
on these important engines will be found in Appendix 11.^ p. 915 ;
Appendix IV., p. 963; and Appendix V.9 p. 999; also p. 1182.

Trials of Boilers, Steam Engines, Gas Engines,
and Oil Engines.—See Appendix I 11.,p. 937 et seq.

Balancing of Engines.—See Appendix IL,p. 897, Appen-
dix IV.) p. 967, and Appendix Ff., p* 1199.

Hot-air Engines.—See Appendix //, p. 915.
Steam Turbines.—See pp. 895, 966, and 1168.




•IP    -***•'


J  >

t      ^     IN*-


**  t r


»    w:   w

« «
» »

I!a4.^ I

fiiii i


^ -S»   sp                     IE

f-rPf  |
*•**•*!.   i

^:TB £

fell if*! 1«


Head, Pressure, and Velocity Energy.


To connect head and velocity:  a water particle of weight zev
while at A, Fig. 694, has a potential energy #/H, and when fallen

to B a kinetic energy of - .   Neglecting friction and other losses,.

and 7'


When water flows steadily between reservoirs kept at constant
level, any portion of water will, neglecting friction and viscosity,
be in possession of an unvarying amount of energy, which may
be due to head, pressure, velocity, or all three. In Fig. 695, a
pressure column A falls short of level c, a portion of the head
energy having become kinetic; and the total head J^ consists

P                                   v*1

of H due to unexpended fall, -= due to pressure, and — due to

G             r                    2g

velocity.    Multiplying each by w gives the respective energy, and
the energy in one Ib. of water

P         7/2

5 + T


An interesting experiment, due to Froude, is given in Fig. 696.
Two tanks, A and B, have discharge pipes c and D, the former
throttled at E, and the latter expanded at F, causing the velocity
energy to become respectively greater or less than at the tank
mouth, as shewn by pressure columns. Further, the horizontal
pressures at E and at F exactly balance, and there is no tendency
to move the pipe.

The Jet Pump. — With sufficient throttling, the pressure
may be reduced below that of the atmosphere, the principle
employed in Prof. Jas. Thomson's jet pump, Fig. 697. Water,.
under a good head, enters pipe D, and passing through the nozzle
at a high velocity, produces a partial vacuum around it. More
water entering at A to fill the gap, the combined streams dis-
charge at B, and thus a field may be drained or other work

Discharge of Water from Orifices. — A tank being
emptied through an orifice near its bottom, the volume of water



*l|f   |fy


if * i»

4l     f     i

» /

*  V-     ,#fmf

^f,^,f|       | t   —                      |*|

*"     ><|i,

t^fl^l-f^   ^    |f
^#!f f^"              fl

<*   I      t

If        if  '    f*       I! >f'"i I

and Coefficients for ditto.


(Fig. 698), and C by gauging actual discharge. Fig. 698 shews
at A a sharp-edged, at B a re-entrant, at c a cylindrical and
external, and at E a bell-mouthed orifice. At B the contraction

	— ~
 _ / j


is greatest by reason of the abrupt deviation of the stream lines;
at c there is contraction within the orifice; and at E no free
contraction, so that there C « c


Gauge Notches.
TABLE OF COEFFICIENTS (average value).

Orifice is C





Measurement of Stream Horse-Power by Gauge
Notches.—Let a stream be partly dammed, the water flowing
through the rectangular notch, abed, Fig. 699. To find the
discharge, divide H into very small portions h, and treat every
small rectangle as a separate orifice, whose area will, when ^ is
infinitely small, be shewn by B, At any depth Ha, # = S^/Hp
l%d discharge through small rectangle = 8 B A/H^. Shewing the
various discharges by horizontal lines on base ef, the figure is a
parabola (the lines a N/H^, whose base is 8 B \/H. Then

Theoretical      ^

cub. ft. per sec.    J

Actual discharge Qa = 5\ C H B JH
where C, the co-efficient of discharge

= '57 4- {breadth of notch -~ (10 x breadth of weir)}

Prof. Jaines Thomson adopted the triangular notch A, where
B/H is constant throughout, suspecting that C would be thereby
regular; and he found that Q a Hf. Taking an apex angle
of 90°,

Qa per sec. == 2-635

where C

Horse Rower I
of Stream   j

•617 (included in coefficient 2*635).   Finally, for any

( available

foot pounds per sec. x 60     Q G
__                  __

„ ) height of

* 1   faUin

I    feet.

Fluid Friction.

The head H is determined by a stake placed in still water
above the notch.    (Seep. 1201.)

Fluid Friction.—The general laws, p. 557, state that
Fnoc #2, and is independent of pressure, but depends directly on
the wetted surface. Measuring the surface area A in square feet,

Fn   =

at moderate speeds, where p = "004 for clean varnished surfaces,
and -009 for a medium sand-paper texture (Froude).

Friction in Pipes is principally due to surface or skin
friction, viscous resistance being extremely slight. Assuming
G = 2g approximately, and placing these values so as to cancel,

Total Fn =

Supposing, now, a piece of water of length L and diameter D
of the pipe, is being pushed through the latter at velocity v:

Fn per sq.ft. of I
sectional area  J      r


~, we divide by G, and obtain

Head lost in friction

/• 965.)



&t l'f|;    J< <, ^/liirn ^j,  >4t

v1  >/ -'^ »l;   ** o^i,  r !»T^|

I'lf-W      I,,      S^r'|^|f <»     ft      ffrl&ttflffiffi,



f'        ,    *},   '    ^     *r*'/ J     ^i      I*     (^   *  *      * ^^

'*       '     fl

-     (  \ j i   ,   >    S4    • j

Itt                 If <


vf* * if|| %i               «

jw *f **v*               (ti I - 4i it

/                           „?

;** i

i^f         |^j*



f        ** I       f|«

t                     |jf

^^^   *,!? 1

. *.'__!

|,» in i H*



4^', i
*  f, /   tft   ,j^l




**    Ji#t0

f^%     f»*jf|    ,*




>1|V              *«

t                *jtt'«


W|                   ff    *»

JT              4                 *||| n  |fe»j it|

-jf ^i '*•*** it                            >«{} w* i

*«f|   p                         Jftjl

f   J^MJp   Jfyl    |^>


'fnitiil*!^* ,

|          M|

| ^mJtl^     f     |*|


i tit                |«

ii i l^jf                |«*   |

| *f^* •%«

C|f wtjj*               Jiff


'   jt »

ll/ri ^   4 lf

>«* |j*l                   *r 4*1

|           «||    | I      »/oiU"*1tyJl     ^I|  |                 Jf||

|,tfn|          -4    >^|   *|f   (%f»^'H|*i4

>  • ,-j^ i-M

/,   f^ "i;

Water-wheel Governor.


decreases, the balls fall and put D in gear, thus turning G oppo-
sitely, and partly opening the sluice. The governor is driven
from the water-wheel by a belt.

The Undershot Wheel is shewn in Fig. 707. The form of
float has been drawn at Fig. 703, and there only remains to add
that, with Poncelet's improvements in floats and race, the water


The Pelton Wheel.

leaves the wheel with little absolute velocity, and the efficiency
is about '66, a great improvement over that of the old radial-float
wheel, which was only '3. As the water never fills the vanes,
there is no pressure, but pure impulse only, and the efficiency is
therefore constant under varying sluices. Horse-power may be
reckoned from head or velocity (see pp. 719 and 720). The
circumferential velocity is about "55 of that due to head, and the
jet thickness is about 8 or 10 ins. The wheel is suitable for falls
up to 6 feet, and the diameter may be four times the fall.

The Pelton Wheel, Fig. 708, is an American machine, in which
a small jet issues from a nozzle A, with great head, and impinges
on a series of cups B B, of the form of a split semicircle in end
elevation c, and simply cup-form in side elevation D. In this
way the jet, about f" diameter, is split, and returned without
serious shock. In one example 320 H. P. was given off from a
fall of 523 ft., the nozzles being one inch diameter. The efficiency
is commonly '8, but may reach *p.                                  " -

The Fourneyron Ttirbine.


Turbines, formerly including only horizontal types, is the
term now applied to all water wheels in which a relative move-
ment of the water to the wheel causes reaction. The Reaction
wheel, Fig. 709, is the earliest form, being a turbine without
guide blades. The casing A, or wheel proper, has tangential
nozzles BBB, through which the water leaves, entering at c; its
reaction on A thereby producing motion. If the best velocity,
that due to head, be employed, an efficiency of '6 is attainable;
but otherwise there is considerable waste of energy. This fact
led to the introduction of guide blades and curved vanes, and the
invention of the true turbine.

Fig. 709.

The Pourneyron Turbine, Fig. 710, is an outward-flow and
also a pressure turbine, the wheel passages being kept full. A, the
wheel, is keyed to shaft B to transmit the power, and the water
flowing downward from c is so deviated by fixed guide blades DD,
that it enters the wheel nearly at a tangent. The wheel vanes are
so curved that the flow is then changed to a radial direction, the


Jonval and Girard Turbines.


water leaving with little absolute tangential velocity, having given
some 70 or 80 % of its energy to the wheel. Regulation by
throttling always reducing the efficiency considerably, the wheel
is divided by horizontal plates at G, so that in the drawing there
are three separate turbines which can be shut off in succession by
lowering the hollow cylinder F. Oil is supplied to the footstep
j through a pipe, but immersed footsteps are now superseded.
Horse-power may be found either by head or impulse formulae.

The Jonval Turbine, like the Fourneyron, is a pressure turbine;
but while the latter works best above tail water, the Jonval is
always drowned or else connected to tail water by a * suction'
tube not more than 30 ft. high, and therefore full of water. Thus
a certain head may be saved, which might be lost, through com-
pulsory position of the turbine. Fig. 711 is a vertical section,
where A is the wheel, B the guide blades, and c the shaft; and
the water flowing parallel to the shaft gives the title * parallel
flow? to this class of turbine. Regulation, formerly effected by
throttling, is now preferably obtained by closing a number of guide
passages, preserving complete admission for the remainder. In
the figure the guide passages form concentric semicircles G G in
plan, and are so bent in elevation as to meet the wheel passages
A A, which form a complete circle in plan. This arrangement
provides retiring room for the sluices F F.

The Girard Turbine was introduced to provide against the
loss of efficiency which always occurs when pressure turbines
work with fractional supply. This fault being due to the
attempted driving with a pressure for which they were not
designed, Girard widened his wheel passages towards the outlet,
and ventilated them so as never to entirely fill them with water.
The energy is then purely due to velocity, and the turbine is
an impulse machine; it has also a parallel flow and complete
admission to whatever guide passages are open. In Fig. 712,
A A are the guide blades and B the wheel. The latter is keyed
to the hollow shaft p F, which, continued upward, joins the
solid shaft G and transmits the power. The whole is hung on a
pivot bearing j carried on the fixed pillar H, and the same
arrangement appears in Fig. 711. The guide passages may be
closed by vertical shutters K K, whose rods are coupled to rollers


Thomsons Turbine.

L L lying in the groove M M ; and as the ring Q is revolved, by
hand or governor, through gear N, the shutters are completely
raised or lowered, according to direction of rotation.

In Fig. 713 the actual path of the water is shewn in a Jonval
turbine at A, and in a Girard turbine at B, a b being free path and
velocity due to guide blades, and b c the wheel velocity; a c is the
relative velocity, and shews actual path in general direction.
Making c d = 'b c, ad will be the line of wheel vane causing
curved water path a <:, the horizontal ordinates of curvature on a d
and a c being equal




	*.* y





 •s  •
	^   fu



Fig. 714 is a diagram shewing comparative efficiencies under
varying openings. Although the Girard is usually less efficient
than pressure turbines with full sluice, its efficiency is unimpaired
by fractional opening,

Thomson's Turbine, Fig. 715.—Here the supply water A enters
the rim of the wheel B, and escapes axially into c the tail race, so
the machine is called an inward-flow turbine. Its energy is
largely due to pressure, the outlet being either drowned or
connected with a suction pipe. Referring to the plan, the guide
blades D D are pivoted at E E, and can be moved in or out by the
levers and links F F. Then the vertical shafts at F F are all
connected, and rotated, through worm gear, by the hand wheel G;
thus more or less water mcty be admitted to the wheel. Although
the gear is complicated, its action is very perfect, the supply being
regulated without materially affecting angle of blades or other
conditions, and a nearly maximum efficiency of 75% obtained for
-11! onenings. The wheel is shewn in detail at H.


f   t     '     f       ^   f   ft

5fc*f«f'|    '"'**   •       j* *

*Wi*»^l    ^«^
^ f* <« »s» *«|f 44 v

J  *  *  '      '    *   <        I    '    »   '* '    ^   t «         4  .

"     f      I            V*   *    '*   *    /

Wjyw*1'  ^#*i^«s *m*v4«f *^</

f                 *Tf *h>*''f               *»    *'«*r

t1?^* ^» '' «*f*

..  . .

,   ...,..•,

;                     **fc>                    «*4


»»U»->*   f   *|

i   'i
it       -


19*                   >>l

^|f,*fiN| f                        f

T*II fi^f* rft.fi
^ 115


*  i^lf                       ^i|i

** **

t*s jypi **!   jt)), ! rf i '$fr   rf»|f ,f J^

'«4^jj%i«i -   ,   %*4pj|»|

p    &^   |'#*»|   I        J%* Jl  M4»

|   s«>t»^   *«'4|i   rf|Mt^|4   ,ii| |4
^f    l^'wrji    ^«*|#Ml4»i¥|

/j^ j i^^|»»> ^41 wjtfif


'^ ' *tft ;4tfi»/* 7T*1!!  *>>*# *i§ 4|l
^t   ^^f^|  h*^^   ***||

^4                                  *        ^l, t

fc^l-"*|*E|    |'^|    **

ff}i/l<** /    '* .•  *   *'f i- " t if ;   i-»v  '? i j ;,' •      i **' ^,*i^i*

u * \

The Impulse Ram.


rises; the sluice is then slowly opened while the shaft is being
rotated, and pumping becomes continuous (scale of figure ^).

The Hydraulic Impulse Ram, invented by Montgolfier,
•enables a large flow of water with small head to lift a smaller
quantity against greater head; and is commonly used to provide

if^II;f*   III f &

ST^S'lSfr-i-s,   £

m^^^^^^^%^^$    *        v*'
**   %    sr   *    —•  *   %    ^    *    "    .^    *"    *            "

3: I!

-*i ^   *  j.. •*-

-* t f I

*fc%         **i:         ««                  "'"rf^

3 I; i! r ; ? > ; - ;; :
ilrjf ?;-r^-- '

%   * *

*    9    »    ^    »    f    *

The Worthington Pump.


one suction valve to each end of the cylinder,* and the plunger
becomes a piston.

All the preceding may be driven by steam power. The oldest
steam pump yet in use is the Cornish engine A, Fig. 726. Its
pumps are of the lift type, arranged in relays with less than 30 ft.
between each pair, and the water is lifted from tank to tank till it
reaches the surface. Two forms of f donkey pump ' are also
shewn at 4, Fig. 447, and i, Fig. 448, pp. 486 and 488, where the
engine valves are operated from a crank shaft There is, how-
ever, another class of pump which dispenses with the crank, being
therefore called ' direct-acting,' and probably the best of this class
are those that necessarily work in pairs, being termed * duplex.'

The Worthington Pump is a duplex steam pump, its ordinary
form being shewn in Figs. 720 and 7200. Two steam cylinders
side by side at A, have pistons connected directly to two pump
plungers at B. When a D valve is employed for an engine
working without expansion, the valve and piston strokes cross
mutually at half phase, and the piston cannot then directly
actuate the valve. In this pump, piston No. i works valve No. 2,
and piston No. 2 moves valve No. i, by lever gear, the motion of
the two pistons being alternate; thus, levers L and M rock valve-
levers / and m respectively. The valves and pistons are, however,
so interdependent, that immediately steam enters either cylinder,
the action of the engine commences as a whole, and will continue
unless special friction difficulties intervene. To enable each
piston stroke to be completed before the valve reopens to steam,
the exhaust ports c c are separate from the steam ports D D ;
a quantity of steam is thus also imprisoned as a cushion. In the
pump, E EX are the suction, and F FX the respective delivery valves,
small and numerous, to give sufficient area while diminishing the
closing blow. The arrangement, also, enables the pump to both
draw and deliver at every stroke, and the contrivance is double-
acting ; in addition, the air vessel j equalises the flow, and the
water leaves at K.

The expansive use of steam has been provided for in the
Worthington high-duty engine, Fig. 721.    The engines are a pair
of tandem-compounds, where A is the high-pressure, and B the
* Except in the case of the accumulator pump, Fig. 722,

3 c


Accumulator Pump.

low-pressure cylinder; and each engine works its neighbour's
valves. Thus the lever c of the opposite engine moves the rods
D and T, from which a system of link work connects to Corliss
valves, the fulcra above and below the cylinders being used re-
spectively for the exhaust valves F F, and steam valves E E. Aftei
use in both cylinders, the steam exhausts into the condenser G,
from which the water and vapour is withdrawn by the air pump H,
and delivered into the hot-well s. The pump itself needs no
description, but special attention must be drawn to the means
by which the driving force is so equalised as to be nearly uniform
when delivered to the plungers. Compensating cylinders L L, or
' pots/ rocking on pipe trunnions, contain water under a steady
pressure of about 200 Ibs. per sq. in., and have plungers pivoted to
the pump rod. This pressure constitutes a resistance to the steam
pressure during the first part, and an assistance during the second
part of the stroke, much in the same manner as the inertia of the
reciprocating parts, and the effect on the work diagram is shewn
at M. a and b are the indicator cards; c and d shew the pressure
exerted by the pot plungers, c assisting, and d opposing the steam
pressure : e is the combined effective-pressure diagrams from both
cylinders; and/is the resultant pressure on the pump-rod after
adding c and deducting d. The pot pressure is kept sensibly con-
stant by the intensifier N, whose larger piston p is under an air
pressure of about 75 Ibs. per sq. in. from the air vessel K, due to
the ^ater column ; and the smaller area Q is exerted on the water
I |                       in the pots. The arrangement constitutes a sort of governor, which

\ \ I                       controls the pump stroke, shortening it if a pipe happens to burst.

|                       To accurately adjust the pot pressure, some air is admitted under

|j                       p by cock R, causing a pressure of about 35 Ibs. per sq. in.   These

pumps are constructed by Messrs. Jas. Simpson & Co.

The Accumulator Pump, Fig. 722, is a double-acting pump,

requiring but one suction and one delivery valve. On account of
the great pressure to be resisted (750 Ibs. per sq. in.), an air
vessel is inadmissible. Referring to Fig. 663, in addition to Fig.
722, the piston A has twice the sectional area of rod B; so when
A. moves rightward, displacing the whole cylinder volume through
delivery valve D, half returns into B, and half goes to delivery J>ipe
b. A, returning leftward, draws a whole volume through suction



Pump Efficiencies.

valve c, none passing D, while the volume in B, or halfa goes to
delivery pipe: thus there is constant delivery, though suction only
occurs on alternate strokes. An additional non-return valve E
permits each pump to be worked separately. (See also pp. 1204
to 1210.)

	fio ^ 70 £
 •AID i

	" — .

	^ •



 10 »«



Pump Efficiencies. — At Fig. 723 is a diagrammatic
statement of the efficiencies of centrifugal and piston pumps
under different heads, shewing that the former are least efficient

TJie Pulsometer.


under large head, and the latter under low head. In con-
sequence, simple centrifugal pumps are only employed for pumping
iarge volumes of water under small head, while positive pumps
are more suitable for pumping small volumes under great pressure.

The Pulsometer is a pump in which steam acts directly
on the water without the intervention of a piston. It is naturally
wasteful in working, but is simple and quickly applied on emer-
gency. Referring to Fig. 724, there are two side chambers A A
to receive the water alternately, and an intermediate vessel H,
whose purpose will be explained. EE are suction and GG delivery
valves, B a foot valve, N the delivery chamber, connected to A by
short pipes FF, and Q the rising main or delivery pipe. To start
the pump, the three vessels are filled through the hole c, the
water resthfg on foot valve B. The ball L being compelled to lie
on one or the other seat at jj, steam is admitted at K, and,
entering, say, the right-hand passage, displaces the water through
F, without agitation, until the level falls to the upper edge of the
orifice. Steam then blows through into F with some violence,
and an instantaneous condensation occurs, causing a partial vacuum
in A. The ball being now drawn to the right-hand seat, water
rises into the right chamber ready for the next stroke, steam
enters the left chamber, and the action is continuously repeated.
The vessel H, though practically uncharged with air, serves the
purpose of an air-vessel, assisting the steady flow into N by the
small head of water which it provides; and to prevent the sudden
shock caused by the rush of suction water, air-cocks D D are placed
on the three vessels, and kept open to a very small amount. The
* Grel' valve at p is often applied to economise the steam supply.
It is simply a short hollow piston, which rises and falls on account
of the difference of pressure within and without it, thus closing pipe
K after a portion of the stroke has been completed. (Seep. 966,)

The Hydraulic Press may be looked on as the seventh
simple machine (see p. 480), and is the basis of the transmissive
principle. Fig. 725 represents the press, with pump attached, as
used to compress cotton bales. The pump A draws water from
the tank B, and forces it, under pressure, to the ram cylinder c,
a rapid exhaust being obtained through the relief valve E when
required* Let D « diameter of ram, and d that of the pump,


Hydraulic Press,

fc". 1


I?, i

while the pump leverage is L : i; from the principle of equal
transmission, one pound per sq. inch on the pump plunger is one
pound-per sq. inch on the ram, and

Total Mechanical Advantage

= Mech. Adv. of press and pump x leverage
area of press      L ___ D2 L
area of pump      i         d%

Neglecting friction.    Taking pump efficiency
= •76, and press efficiency = '95 • both combined = 76 x '95 = "72

(Seep. 1210.)

The ram cylinder should be approximately hemispherical (see
p. 68), and its strength is found at p. 399. The leather collar E is
a most efficient packing, being distended by the pressure water
and pressed against the ram surface. The hydraulic jack, p. 206,
is simply a miniature press, where G is the ram and D the plunger.
Its efficiency is, of course, much higher than that of other jacks.

The Hydraulic Accumulator is probably the most im-
portant adjunct in hydraulic transmission, constituting an arti-
ficial head, in which the water pressure is caused by other material
than water. In Fig. 726, a series of weights at c hang from the

Hydraulic A ccum u lators.


T-head E, and, through ram D, exert pressure on the water within
A B. The weights being raised to position F, are a store of
potential energy, which may be given out at will through the
pipe B. Water is pumped in at A to raise the ram, by an engine
such as that in Figs. 663-5,/. 682, and the latter is automatically
stopped and started from the accumulator, as required, by the
levers at G and H, struck by the load. The pressure water

drawn at B may now be applied to the driving of machines doing
intermittent work, such as

1.  Cranes upon dock wharves, &c.

2.  Boiler-shop and shipyard tools.

3.  Lifts for hotels, &c.              ,

4.  Swing and other movable bridges.

5.  Manipulation of heavy guns.;

In all these cases the pumping-engine will have sufficient time
between shifts to catch up on the machines, and thus a com-


Balanced Hydraulic Lift.

paratively small engine, working all the time, may serve for very
heavy work occupying only a short period (see Case 4 especially).
It is in the great storing capacity, and the little loss (skin friction
being independent of pressure, and water incompressible) that
hydraulic transmission is of such immense advantage. The usual
large pressure, 750 Ibs. per sq. in., is adopted because the friction
is then much less in proportion to power transmitted, area of pipe
being small. Chapter VII. illustrates hydraulic transmission
applied to Case 2, and the student may now refer to pp. 292-3,
301-2, 314, 317, 320, and to Plates XV. and XVI., also to
Case 12, p. 580.

Fig. 727 shews Mr. TweddelPs Differential Accumulator,
where great pressure is obtained by considerably decreasing the
ram area. B is the load, and the effective area of ram is A
minus a. Comparing with Fig. 726, it must be understood that,
weights being equal, we lose in time what we gain in pressure, and
thus this apparatus is specially suitable for small machines, such
as portable riveters. The work stored in any accumulator is the
weight or load, in Ibs. x the height lifted, infect, or
w H foot pounds.

A Hydraulic Lift, as devised by Mr. Ellington, and known
as a 'balanced' lift, is shewn in. Fig. 728. A long ram A,
working in a cylinder c, thereby lifts a cage B, and the load .
consists of (i) the cage, (2) the people or goods, and (3) the
ram weight, the last., two being variable. In the older and
dangerous method the average load was balanced by a weight
hung from a cord carried over a pulley, and connected to the top
of the cage; but here the cage and people are lifted by separate
water columns, while the varying ram weight is supported by a
head which similarly varies. The variation in ram weight is due
to the ram's varying immersion, the upward support from the
water (apart from artificial pressure) being equal to the weight of
fluid displaced. Referring to Fig. 728, the pressure from the
main is led to the cylinders r> and E. Upon piston F is a constant
pressure, through L, supporting weight of cage + ram when down ;
and on piston G, through K, pressure water is admitted when
required, supporting the people + friction, viz., the nett load.
Both these pressures are used to intensify the water in M, which



is directly connected to the ram, and on account of such intensi-
fication the ram diameter can be as small as we please, merely
strong enough to prevent its bending. The volume in M being
just sufficient to fill the ram cylinder during full stroke, the pistons
F and G fall to the bottom of their cylinders, due not only to
pressure from main, but to a constantly increasing weight of water,
It is this weight, due to water, filling nearly the cylinders D and E,
which, bearing on pistons F and G, so intensifies the pressure in M,
as to support the whole unimmersed ram weight; being clearly a
maximum when the cage is fully raised, and nothing when the
cage is lowered to the bottom. The varying ram weight is, there-
fore, correctly balanced in all positions, and the only load to be
averaged is that of the people. When lowering, water is exhausted
from N, and the descent caused by the weight of the people.

The cord P, passing round a pulley on the working valve Q,
will open the latter to pressure p, or exhaust ey in any position of
cage. If the water in M decrease through leakage, the cage is
lowered to the bottom, and water at N exhausted : then pressure
water being admitted at R, the pistons are forced upward, com-
pelling some water to pass from above to below piston F, through
its packing; at other times R is empty. (See Appendix V.,p. 100,6.)
Intensifiers, or intensifying accumulators, are a means of
transforming small pressure, as from a town main, into a really
useful hydraulic pressure. Recent descriptions will explain the
principle, and good examples will be found at pp. 375 and 836.

Hydraulic Cranes have many advantages over others.
Being worked intermittently, a small pumping engine will store
the power: the.latter, again, being used with considerable rapidity
and saving of time, a consideration when loading vessels at wharves.
The lifting, too, being done without vibration or noise, makes
these cranes of special use in raising foundry boxes and other
like work. The cranes are also very simple.

Fig. 729 shews a cylinder, ram, and pulleys, the essential
apparatus for each motion of a hydraulic crane. Cylinder A has-
a common stuffing box c, packed with hemp, and carries a number
of c fixed' pullies, DX D2 r>s, the ram P supporting an equal
number of * movable' pullies EX E2 E3. To prevent the ram
turning on its axis, the head F slides on guides G G, and the

Hydraulic Cranes.


whole apparatus is fixed to the crane by feet j j. A wire rope or
chain being attached to the eyebolt K, and carried round the
pullies E1D1E2D2E3, leaves D3, by w, to the load or slewing:
wheel, as desired. Examining by the pulley principle Fig. 439,
p. 483, the mechanical advantage will be inversely as the number
of cords or chains at L L, Fig. 729, p being now the greater, and
w the lesser force. Neglecting friction,


P "~ no. of cords

Mech. Adv. = —

And allowing for all resistances, •



no. of cords

where  efficiency 77 varies  with  the number of pullies,   by the
following table, found from practice :



77 =

and the greater tension, at tail end, equals P •»» (no. of cords x 17)*
Thus, a heavy pressure with slow speed has lifted a smaller
load at greater speed, the distance between pulley centres having
been increased.

In order that the ram shall finish its stroke quietly, automatic
cut-off gear is supplied. Valve H being opened to pressure by
raising rod N fully, the ram, ascending, strikes a tappet R by means
of the projection Q, when the stroke is nearly complete, thus
causing lever M to be pulled over to position s, and closing the
valve. A further movement of M to position T opens H to
exhaust, and the ram descends by the pull of the load,

Reference may now be made to Plates XV. and XVLy
shewing various hydraulic cranes. That on the left in Plate XV.
is the best example of pulley gear. Thus, cylinder D is for
lifting, E for traversing, and c for slewing, all worked from valves
at s.

742                            Working Valve.                                                  \

Working Valve.—When a D slide is used, Fig. 730 is the                  f

usual form, where p, R, and E are the passages from pressure, to
ram, and to exhaust, respectively. At B the valve is. open to
pressure, and at A to exhaust, while at c the ram passage is
entirely cut off, by hand or automatic gear.                                                «

Hydraulic-pressure Engines, though wasteful with small                   f

pressures and high speeds, may reasonably be used when supplied
with water at 750 Ibs. pressure or more, the piston speed being
not more than 80 ft. per minute.     The first piston engine, in-
vented by the late Lord (then Mr.) Armstrong in 1838, was of the                  !
rotary type.    Subsequently he adopted side-by-side cylinders with
reciprocating pistons, and in the present engine, as applied to                   ^pr
heavy work, such as turning ships' turrets or swing bridges, there                  i
are three oscillating cylinders, whose pistons connect to the same                   I
triple-throw crank shaft, and each valve is worked by a rocking                   \
lever on the trunnion.     Fig. 731 is a section through one valve                     j
box.    Valve A is reciprocated by the trunnion lever, while valve B,                     ,
used for reversing purposes, may for the present be considered
fixed,     c is the pressure supply, D the exhaust pipe, and E F the
connection to the cylinder.   Taking present position of B, a right-
hand movement of A admits pressure to E, and a leftward move-
ment permits exhaust from E, through H and G, to D.    Supposing,
now*, B'S position be so changed that H is opposite D, and G                   \
opposite F ; the conditions are reversed, and a leftward movement                   \
of A admits pressure to F, while a rightward movement exhausts                  ^
through the valve to D.   Thus B is a reversing valve, and is moved
by the piston of an auxiliary cylinder.                                                                   1

The Relief Valve j is simply a small, spring-loaded safety
valve, which permits an escape of water whenever the pressure
exceeds the normal, by reason of water inertia Such valves are
placed wherever there is liability to shock.

The Brotherhood-Hastie hydraulic engine, Fig, 732, is a com-                   \

bination of the well-known Brotherhood engine, p. 632,  with                  y

Hastie's automatic stroke adjustment.    Pressure water entering                  £

at P, passes to the cylinder by pipe A, and the exhaust returns
through the same pipe, but is diverted by valve D into the outlet E.                 /

If p and E are connected to a reversing valve, the pressure water                  |

may enter at E and leave at p, and the direction of engine rotation                  i


Hydraulic Pressure Engine:

is then reversed. The principal feature in this engine is the
.crank pin B, which is not fixed, but capable of sliding to a limited
.amount, within a diametral groove in the crank plate c, being for
this purpose screwed into a shoe plate M. The power given to c
is transmitted by a hollow shaft Q, through the strong volute
spring H, to the driving pulley u. Now u is keyed to the inner
•shaft K, and when the load conies on the pulley there is a further
•coiling of spring H, which causes shaft K to turn relatively to G,

through an angle depending on the turning moment. The result
•of K'S turn is to rotate a cam F in such a way as to move the
•crank pin further from the shaft centre, and thus increase the
throw; while on the other hand a decrease in load reverses the
cam movement and enables the piston pressure to shorten the
crank centres. Now there are two ways of accommodating fluid
pressure to work required : alteration of stroke or of pressure. In
the steam engine reduced work is met by reduced pressure; but
•water, being inexpansible, can only be adjusted in supply by a


with Variable Stroke.


corresponding adjustment of stroke.    The result is roughly the
same, for pressure x stroke = work done.

The cam F is peculiar in shape; it is shewn under full load,
having turned through three-quarters of a revolution, in a right-
hand direction. Its highest point is'at-f-, and its lowest at #,
and when the load is removed, the cam turns leftward until the
projection at * stops itself against the projection G. Similarly
the earn shewn dotted serves when the engine rotation is reversed,
the projection j being then acted upon. Both G and j are one
with the shoe-plate M, but lie in different planes, so that the two
cams, also in one piece, may rotate without interference.





P. j/. Brass and Gun-metal Founding.—With these
alloys a good height of pouring head is required, so as to cause
pressure and prevent porosity ia the casting. An ample number
of risers should also be provided to permit escape of air or gas,
and thus avoid honey-combing. No blackening is used, but the
mould is faced with very fine sand.

JP. 42. Steel Casting.—It having been noted that small
castings were more porous than large ones, the conclusion was
arrived at that it was of the utmost importance to keep the metal
at a very great heat until poured. The present method of pre-
venting blow-holes is to add silica-Spiegel (a combination of
silicon and pure cast iron) to the ladle while pouring, the better
plan being to apply it ia the molten condition. If this substance
be not thoroughly mixed we get 'hard' and 'soft' spots in the
casting, the former being due to accumulations of the silicon.
In order to keep the casting uniformly hot until the whole mould
be filled, the pouring gate should be chosen at the heaviest por-
tion of the casting; and small castings are preferably poured from
a small converfer to themselves, instead of from the refuse of a
large open-hearth melting. By facing the mould, where required,
witli ferro-raanganese or ferro-chromium, very hard surfaces are
obtained in those places, the latter substance giving the hardest


Appendix /,

P. 85.  Phosphor  Bronze. — The proportions by weight
are as follows : — .

Copper     ...    85

For ordinary uses


f   Cc

]   Ti

(   PI

r  cc

]   Ti

(   Pr




Copper     ...    90
Tin          ...      9

Phosphorus         "75


Tough metal for piston rings
and eccentric linings

Phosphor-tin      7


For bearings (heavy machinery){   ^5]£or-tin   ?s



V P. 102. Welding. — It has been shewn by Sir Thomas
Wrightson that the phenomenon of welding is akin to that
of regelation^ or the sticking together of ice under pressure. To
prove this an experiment was made which shewed a distinct
decrease in temperature (amounting to 106° F at 2550° F.) during
welding, a similar result being known to take place during rege-
lation. This abstraction of heat is caused by the melting of the
iron or ice in either case, and the consequent need of latent heat
for the liquefaction.

< P. 124. Case Hardening.—If two pieces of iron, forming
pin aad socket respectively, are to be case-hardened, and a good
working fit be finally required, it is important that the pin be
made a pretty tight fit in the socket before hardening. After the

Appendix I.


hardening process, both pieces will have swelled in volume,
and it will be found that the pin will fit the socket more tightly
than before. The final fit may be obtained by lapping the pin
with emery powder in the lathe.

1 P. 128. Hardening Steel.—The difficulty experienced in
hardening milling cutters without cracking is found to be largely
due to unequal heating as well as unequal cooling. To avoid the
former a method of heating in a bath of molten lead kept at a
high temperature is found to be very successful. Regular cooling
is very difficult to obtain in the case of thin articles, such as
circular saws; but by placing a sheet of brown paper upon the
surface of oil and allowing the article, placed upon the paper,
to gradually sink into the liquid, warping may be largely pre-
vented, though nothing softer than the equivalent of a brown
colour can be thus obtained. If a saw is to be tempered* to blue,
the usual course of water tempering must be followed, dipping
as smartly as possible, and the blade be straightened afterwards.
Hardening in water and tempering subsequently in oil will pro-
duce a softer result than if water be used throughout.


P. 152. Lathe Centreing.—By adopting a very slightly
more acute apex angle for the centreing drill than for the lathe
•centre, the necessity for drilling the small hole, as mentioned at
top of p. 152, is avoided. See Fig. 733.


P- *53> Double Driving.—By allowing both ends of the
•carrier to be driven from the catch plate, stress is taken off the
lathe centre, and more steady tooling is produced. Clements'
driver, Fig. 734, is designed to effect this purpose. The carrier c


Appendix L

is driven by both the pins D D on opposite sides of it, and these
pins are not fastened directly to the catch plate A, but to a
separate cross-shaped plate E. The pins B B, holding plate E to
the plate A, pass through slots in E, so that the latter is permitted
to adjust itself to inequalities in the contour of the carrier c.

lJjg. 734.

P. i$j. Lathe Tools.—A simple and excellent roughing
tool for a lathe is shewn in Fig. 735. A groove being fullered
by the smith at a short distance from the tool point, the upper
surface is ground to suit the tool angle in the direction shewn by
the oblique arrow, while the relief angle is obtained by grinding,
both'at front and side as shewn in the end view. (Seep. 976.)



P. 214. Gauges.—A very handy form of gauge is shewn in-
Fig- 736. It combines in one tool the equivalent of plug gauge B,
and ring gauge A. It is not quite so perfect in its application as,
the cylindrical gauges, but will serve most ordinary purposes.

Appendix I.


P. 230. Turning Eccentric Sheave.—Where much of
this work is to be done, and where the eccentricity of the sheaves
to be turned varies much, the mandrel shewn at Fig. 737 is an
ingenious and useful appliance. B is an expanding mandrel,
having a cone A and four inclined keys D D that can be advanced
outward or inward by the nut G, so as to grip the sheave firmly.
The eccentricity may be adjusted by applying the handle H to
the screw at B, and by unloosing and refastening the nut F. The
whole is placed in the lathe by bolting the frame c to the face
plate j, then advancing the poppet head to the centre F.

. 737.

P. 232. Boring Eccentric Straps.—Instead of turning
the straps in the lathe a heavy drilling machine may be con-
veniently adapted for boring, purposes, as in Fig. 738. A large
cutter c is placed horizontally through a slot in the boring bar F
(or drilling machine spindle), and the radial feed may be obtained
by giving a slight turn to the pinion D, which fits into teeth upon
the cutter. B shews the eccentric straps, which are firmly bolted
to the table A.

P. 232. Planing Eccentric Straps.—The method shewn
at EJ, Fig. 245, may be varied by using a shaping machine and

Appendix I.
v^ork to  the side of the table, as at E,& Fig. 260,

- *  r

* A    1

' Y        Sl^rf;aces must be first shaped.

at  p.

i **"li.*-*htl.


"*?UlS-  awcl 24.3.  Machining Brasses,—There are three
^machining brasses for bearings and rods : (i) where
brasses are cast  in  one,  then bored, turned, and
finally cut through the middle with a parting tool;
1>~er l^ft by the tool is to be filled with a liner of the proper
\2)    The half brasses being cast separately are united
and after tooling are separated by heating: here
(3) Bolting together as-

Efcoring Crosshead.—In case there should not be
large enough for the purpose shewn in Fig. 254, the
may  be bolted to an "angle plate placed on the face
tlie setting is not quite so satisfactory.

* *  A^p»,   Milling a Radius Link,—The appliance shewn
739   is   for the'purpose of guiding, the link under  the
fcool   in   such a manner as to cause a slot to be milled
tlie; correct curvature.    A and B are two slides hinged at c,
;as   sh.ewri (plan view) upon the milling machine table.
a.ire now bolted down so as to> enclose a very wide angle
-* * i%   Ok  lairge   curve is the result, and if the angle be decreased,
it                       curve is traced on the link.    EE are two dies having

whtlcli   carry the upper plate D, and on the latter the radius
is  !>olted. in the position shewn.    The right-hand die can be
by trie screw F, this advance being the feed, and a very
I ho light will shew that the milling tool will cut out a curved
radius will depend upon the angle ACB.    (£eej>. 819.) .

..,MH-""''"''"'*&'  _

Appendix /.


P. 256. Turning Governor Balls.—The method already
described is adopted only where the special appliance shewn in
Fig. 740 is not readily obtainable. The slide D, with its tool
holder, is known as a ball-rest. It is pivoted at c, directly under
the centre of the ball to be turned, and is supported upon the
saddle B. The tool being set to turn the ball of correct diameter,
the necessary radial feed is given by hand. The feed is more
regular if spur teeth be cast upon the curve E, into which a pinion
on the rest engages. A horizontal hand wheel fits then on the
pinion spindle.

. 74O.

P. 256. Cutting Bevel-wheel Teeth.—If not too small
these wheels may be correctly cut by the machine in Fig. 741.
The bars r> D, carrying sliding toolboxes c c, are centred at A on a
universal joint, and as the tools reciprocate a slow conical feed is
obtained by guiding the bars round the plate on ' form ' E, which
is cut to twice the tooth scale and set at twice the distance from A.
(Seepp. 821 ^986.)


Appendix I.

Cutting Key-ways in Wheel Bosses.—Usually the
wheel is for this purpose laid horizontally on the table of a
slotting machine, as in Fig. 742 ; and if taper be required, the
boss is set at an inclination of i in 64 by packing at D. The
clamps c and bolts D then hold the work securely. The cutting
tool is sketched at F. It is now the custom in some shops to
make the key-ways and keys perfectly parallel, but a very good fit
with each other : a bursting pressure is thus avoided.



\    i


* P. 289. Hydraulic v. Electrical Transmission.— In
the discussion on the President's address before the N. E. Coast
Institution of Engineers and Shipbuilders, October i6th, 1894,
Mr. Tweddell says : ' Any remarks on the economy due to
hydraulic transmission apply with equal force to electrical trans-
mission. As the laws affecting both systems are almost identical,
the question resolves itself into a matter of suitability for certain
tasks, and as hydraulic pressure is suitable for intermittent work
or for rectilinear motions, and not so suitable for rotary motions,
it follows that a combination of the two systems .... is exactly
what is wanted.'

¥ P. 328. Electric Welding.—An interesting paper was read
Jbefore the same Institution on February i2th, 1895, by Mr. Henry
Foster, in which he described the methods adopted at the New-
burn steel works. The f Benardos' process was used, in principle

Appendix I.


as shewn at p. 329, and the work to be welded consisted of
general machinery repairs (especially for-engine breakdowns) and
boiler repairs. In addition the arc was used for boring holes in
plates, or for otherwise melting portions of metal away. The
process is of great advantage for breakdowns, as putting the
machinery in working order in an extremely short space of time,
and is also especially useful for patching purposes, thus saving
many articles from the scrap heap. The average ratio of weld to
solid was 85-5 per cent, for iron, and 80*8 for steel, as shewn by
testing. The best hand welds were found to be much below this.

P. 693. Fusible Plugs made of gun-metal are screwed into
the crown of boiler fire-boxes or furnaces. They are drilled
through the centre and filled with a fusible metal whose com-
position depends on the temperature at which it is desired the
steam shall blow out the fire when shortness of water occurs.
The following table will shew the composition required:—

Melting Temperature
in degrees F.




Composition of Metal in parts

by Weight.

Lead.           Tin.         Bismuth.

I                I                4








~                o

(See also App. //., p. 833.)

P. 363-6. Nature of Shear Stress.

Fc - ,/lFg   and   Ft »

But Fc and Ft each act on areas of */2 while F8 acts on an
area = i (see Mg. 324):


whence by substitution,

Apfendix I.
Ft = /, J*

or, a shear stress produces two normal stresses, tensile and corn-
pressive respectively, each of an equal intensity with its own, and
having directions at 45° to the original stress,

•^ 3&5-   Influence   of  Time   on   the   Stress-strain
Diagram.—The diagram in Fig. 743, taken by Professor Ewing,


of Tjurve

cf. 743.


shews very clearly the different plastic lines obtained, according
to whether the test experiment be made very quickly or very
slowly. It is clearly important that an average rate be maintained
in applying the straining load.

P. jgo. Wohler's Experiments.—To give a further
interest to Wohler's important experiments, the three machines
which he used are shewn in Figs. 744, 745, and 746. In the
first an axle is loaded by a spring at a considerable distance
* over-neck/ and is then rotated several millions of times before
breaking, the case being that of alternate stresses. The second
machine, Fig. 745, represents the bending of a beam, The load
is again caused by a spring in tension, and the varying stress is
obtained by the rotation of the lever B, placed below. This is the
case of a live load (removed and replaced). The third drawing,

Appendix I.


Fig. 746, shews a tension experiment.    In every case A is the
specimen, B the rotating shaft, and c the load spring,

P. 399. Thick Cylinders. — The stresses within the
material of a thick hollow cyMnder *are really somewhat com-
plicated. They always consist of (i) a radial pressure, greatest
at the inner circumference and decreasing to nothing at the
outer circumference; and (2) a hoop tension, which is greatest
at the inner ring and least at the outer ring, in the manner
shewn in Fig. 352. This diagram shews what the hoop stress
would be in an initially unstrained cylinder.

Now, in order to shew how changes in mechanical construction
can decrease the hoop stress, and therefore decrease the thickness
of cylinder necessary, we will put out of the question the radial corn-
pressive stress and the possible longitudinal tension, and consider
only the hoop stress as tending to break the cylinder. Just as AB,
Fig, 747, is the* diagram of tensile hoop stress, c D is that for the
fluid pressure, or load^ and as A B = c D for conditions of strength.

as with thin cylinders, only that/ is the average hoop stress,

Now, in this diagram it is evident that only the inner rings are
of much value in resisting the load, and the outside rings do not


Appendix: I.

do anything like their share of the "work, so to speak. We pass,
therefore, to the practical principle of

Initial Stressing (during manufacture), by which the outer
rings may be made of greater. assistance. To do this, they are to
be stressed initially in tension, and the inner rings in compression,
and it will be shewn that when the fluid pressure is admitted, the
internal hoop compression will be more than relieved, while the
external hoop tension \vjll be hut slightly increased.

The first and most important method of arriving at the above
result is to build the cylinder in separate concentric tubes, to be
shrunk, the outer over the inner ones.

The tube B, Fig. 748, being shrunk over A, exerts a pressure
between their common surfaces, as shewn on the left at /$, and
this pressure decreases to nothing at the inside and outside of
the cylinder. But such pressure may be likened to a fluid pres-
sure on the inside of B and the outside of A, and it will be easily

Appendix I.


seen that hoop-stress diagrams like c -and D will be formed, which
can be calculated if we know the pressure between the surfaces.
It must be noted also that c is compressive and D is tensile hoop
stress. When, however, the fluid pressure is admitted to the
cylinder, it will tend to equalise the stresses, for, supposing E
to be the diagram obtained with a cylinder not initially strained,
the diagram E must, in the actual case supposed, be superposed
on c and D, having regard to sign, or be set up on the base line
abed. The dotted areas will be the final result, and the real
point for us to notice is that, instead of stressing our cylinder to
eft it is only stressed to gh, the outer rings taking their share,
and thus the thickness of the cylinder may be much less than if it
had not been initially strained. The areas c and D will be exactly
equal, because there is equilibrium at first.

A still more equable stress may be obtained (and consequently
less thickness required) by adopting a greater number of rings.

Fig. 749 is a diagram of the stresses when three rings are used,
and will be easily understood from the last diagram. As before,
A = B in area.

In applying these principles to the manufacture of large guns,
botl| the initial and maximum firing tensions are kept within
18 tons per square inch. By using the modulus of elasticity it
will be quite easy to find the hoop stress produced at the ring
between the tubes, and conversely the radial pressure there,
caused by extension due to shrinkage.

When guns are constructed by winding wire very tightly round
a thin core, a perfectly .equable stress may be obtained, and con-
sequently these guns are lightest of all. In Fig. 750, dotted lines
shew tension put in the wire as it is being wound on. The curved
line abc shews resulting stresses in the wire after the gun is-
finished, and the thick line the hoop stresses when the gun is

Cast Iron Cylinders, if cast without any precaution, will
be in a state of compression on the outside, after cooling, and of
tension on the inside. Building then the hoop-stress diagram,
Fig. 751, upon the incline, we get a very mucji worse result than
before; for the initial stresses, caused by the inside codling last

'$. 69), only assist the destruction of the cylinder when the



Appendix: J.

fluid pressure is admitted. We therefore require an inordinate

But if we cause the core, during casting, to lose its heat more
rapidly than does the exterior, by circulating cold water through
a central pipe, the inside will cool first, and a tension will be
caused on the outside while a compressive hoop stress is felt on
the inner rings.

Building now the (fluid7 hoop stress on the new inclined line,
we obtain the dotted curve shewn in Fig. 752 The hoop stress
is generally more equalised, and the internal hoop stress (or
maximum) is decreased by the value of the initial compression.
Again, therefore, we have been enabled to decrease the stress,
and consequently the necessary thickness^ by the artifice of initial
stressing, caused by casting with a cold-water core.

Building-up from separate tubes, or by tube and wire, must,
however, be looked on as the very best methods, where, in fact,
an almost certain desired result maybe obtained. (See App. JL,
p. 841.)

P. 4.07. Pitch of Riveting.—It should be distinctly under-
stood that the formula, pitch = 1^09 -t-^i is only true if we use
steel both for rivets and plates, and that we do not adopt plates
above an in,ch in thickness. All other cases must be referred to
the general formula on p. 407. The diameter of the rivet for a
particular plate has, up to the present time, been fixed by practical
considerations of punching. It is open to question, now that
boilers are drilled, whether some alteration may not with advantage
be made, tending to increase the diameter of the rivet for thin
plates. (See Appendix: ///,/. 921.)

JPj>. 442-6. Position of Supports for Least Bending
Moment.—Let the beam A B, Fig. 753, be loaded uniformly,
and let it be required to find the position of the two supports
when the least bending moment shall act upon the beam. This
will occur when the maximum stress caused by the load between
c and D is equal to that caused by the loads at A. c and
D B. Thus, if we considered only the loads at A c and D B, a
uniform bending moment is caused by them between c and D,
and this moment has to be completely balanced by that at the

Appendix L


centre of the beam, caused by the load en. In short, as B^
oc/ oc S, M! must equal M2, in order to cause the least stress
in the beam. Assuming the lettered dimensions to represent


BO,^H       Fjuof.753.



feet, and the load to   be i Ib.  per   foot  run, then the  bending
moments are as follows:—

(i)    Due to distributed load at b = —

(2;    Due to distributed load at a - x =---------

(3 )    Due to concentrated load at E = x(a - K)
(See also Fig. 400.)

(2 x}2     x'2
(4)    Due to distributed load at 2 x = ^-5— = :™

But (24-3) = Mj              and (i) = M1

also (4) = M2

.-.    (4)= (i),

and kv=

Again    (4) == (2 -f 3)

(a - sc

a - oc)


or    2 b^ = dP
and   b~

I H-

».= = '414 /




x /.

There aie two practical applications of this problem to which
reference may be made. The first is where a locomotive boiler-
barrel is sometimes flattened to save width, and the flat portion
becomes such a beam as we have discussed, the supports being
then represented by stay "bolts. The other example is that of a
paddle-wheel float, when supported by the float arms at two
places only.

An extension of the problem is shewn in Fig, 754. There
are two supports, A and E, but the former is a hinge. Using
similar letters,

oc = by        b ='4 1 4/,        /=L-#=L— b
^,           and    £

A similar case, but with concentrated load at c, is that of the
wall crane, Fig. 754.

P. 445. Continuous Beams. — The following is a simple
method of finding the reactions on the supports in the case of
a beam over two spans.

Suppose the mid-support be removed, a deflection will be
caused by the uniform load, 2 Wlt

48 El

(See pp. 451 and 849.)

Now the upward pressure that would neutralise this deflection
would have the same value as a concentrated load capable of
causing it. Let this load = R: then the deflection caused by R,

A<> =.........o -r->T     and equating the two values,

40 E* *•



48 El ~   4&EI

whence E. = —

Appendix L                               763


P. 481. The Lever-loaded Safety Valve. The effect
of the weights of lever and valve may be taken into account as
a downward force at the valve centre, which has to be overcome
by the steam pressure. In addition to the lengths shewn in
Fig. 438 (measured in inches), and the weight w (in Ibs.), Let

w = weight of lever, in Ibs.

v = weight of valve and valve centre, in Ibs,

/= distance from fulcrum to centre of gravity of lever, in inches.

/ = pressure per square inch of the steam.

d= diameter of valve, in inches.

Then,        Upward moments    =    downward moments


P. 487. The Quadric Chain. Referring to Fig. 449,
the examples at 3 and 4 are spoken of as parallel-crank chains ;
while that at i is called a lever-crank chain, from the fact that
the beam rocks like a "simple lever, and the crank completely

P. 503. Oldham's Coupling, Fig. 474, is derived from
the elliptic trammels, Fig. 448, where the trammel bar is fixed and
the cross revolves. The relative motion being unchanged, a fixed
pencil held against C's face will produce an elliptical arc. Also
C's centre will describe a circle passing through centres A and B
(Euclid, III. 31), for the angle between grooves is always a right
angle. Hence (Euclid, I. 32) the angular movements of A and B
are identical.

P. 312. Approach and Recess. During approach the
flank of the driving wheel is acting on the point of the driven
wheel (see Fig. 501), while during recess the point of the driver
acts upon the flank of the driven wheel. On the supposition that
the pushing friction during approach is more prejudicial than the


Appendix I.

trailing friction during recess, some clock makers cut away the
flanks of the driving wheel and the points of the driven wheel.
A much larger number of teeth are then, however, necessary to
secure smooth action.


<                   .                       i

Pp. 594. and 620. The Dryness of Steam.—The pro-
portionate dryness of steam at all points of the stroke can be
very conveniently represented upon the indicator diagram by
means of a simple construction. Imagine the indicator cards in
Fig. 755 to have' been obtained in the usual manner, and then to
have been plotted by the method shewn on p. 622, the clearance
volumes being of course known in terms of the cylinder volumes.
Let it also be supposed that the weight of steam passing through
the cylinders at every stroke has been found experimentally by
measuring the feed water entering the boiler, the latter being of
course only occupied in supplying stearn to the engine. Referring

a. 755. •

next to the diagram on p. 598 or to the table of saturated steam
volumes in this appendix, • the volume of i Ib. weight at 70 Ibs.
absolute pressure is found to be 6*1 cub. ft, from which the
volume of steam due to the feed water weight can be at once
deduced. Similarly the same quantity of steam, when expanded
to 30 Ibs. absolute pressure, will have a volume that may be
ascertained from the tables, or can be found by constructing the
saturation curve,

Referring again to the diagram* the compression curves are to
fee continued, as at A.B and CD.    Then at B E set off true steam

Appendix I,

volume at 70 Ibs. pressure, and at JD F that due to 30 Ibs. pressure,
Through E and F respectively dra\v the saturation curves E G and
F H, which for rough purposes may be hyperbolas with origin o;
but, for greater accuracy, may be drawn with a new origin for
each curve, according to the inarmer described on p. 624, and
shewn in Fig. 624. Thus JE being divided into 6'i parts, and
K F into 13*5 parts to find the scale of specific volume, the origin
for each curve must be moved "41 ft. to the right and '35 lb.
below the old origin, and the curves are then treated as
hyperbolas from the new origins. The method of constructing an
hyperbola is shewn at p. 615.

It will be seen that there is a separate saturation curve for
each cylinder of the compound engine, but this is only due to the
fact that the clearance volume in the L.P. cylinder does not bear
the same relation to the steam volume as did that in the H.P.
cylinder. If it be desired to shew both cards under one
saturation curve, it is only necessary to transfer the lower card
to tracing paper and move it to the right until the curves E G and
F H coincide. Or another method is to re-set out the cards by
placing the curved lines c N D and A M B upon the vertical line
o K j, and so move the cards to the left, distorting them some-
what. One saturation curve may then-be drawn through E, B E
being the same as before, for the clearance steam is now
eliminated. Of course, instead of completing the compression
curves by the dotted lines, the curves could have been drawn for
any pressures between A M in the H.P, cylinder, and between c N
in the L.P. cylinder.

Assuming now that the cushion steam is dry—^n,, assumption
involving but very slight error—the dry ness of the steam may be
found for any point in the stroke of either piston, fcy comparing
the total volume of steam in the^ cylinder with that shewn by the
saturation curve at that point, the difference of these volumes
shewing the amount of steam condensed by the cold cylinder
walls. Thus, at commencement of the H.P. stroke:—

Dryness fraction =» —


Wetness fraction = —
j E


Appendix T.


And these quantities may be ascertained for other points of the
stroke by measuring similar intercepts along horizontal lines.

It will be noticed that the steam always tends to become
drier toward the end of the stroke (due to re-evaporation), though
never becoming entirely so ; and the value of compounding in
drying the stearn is also apparent. (See Appendix //.,/• 878.)

P. 608. Adiabatic of Saturated Steam.—Zeuner gives
the following empirical formula to find the value of the exponent
n in the equation to the adiabatic for saturated steam :—

#= i'035-f ("i x dryness fraction).

from which it would appear that Rankine's curve is suitable for
steam having 25 percent, of suspended moisture, while Zeuner's
curve is for stearn initially dry.

P. 6op. Expansion of Wet Saturated Stearn.—It has
been already pointed out that the saturation curve is a curve of
lowering temperature. Also that the compression of dry satu-
rated stearn at constant temperature causes it to become wet by
partial condensation, because steam at a given temperature can
only exist at a certain pressure. Conversely, the expansion of
stearn at constant temperature will tend to dry it; but if there be
sufficient water present, the pressure will not fall during the ex-
pansion, and therefore the curve on the diagram would be a
horizontal straight line.

Calculation of Specific Volume of Dry Saturated
Steam, when the temperature, pressure, and latent heat are

. 756.

Dealing with i Ib. weight of steam, whose volume and pressure
are shewn at A, Fig. 756 :    •        •         •


jftt   <t«*t* f

*  Ifi     / ' ///                                 M\  ^      V*"'    ^   *'    ** '"'-'> '

**!*•**»'* «•$ $**ij, JJM  ^#j ^  r^)*,  a!"^*   f     *'   .***^*'f«»f **

^      >*>f   '|l"    «   ;*   # V         *   -  fJl**   ^'      H     f   |^

^       II   >! *  I**   4/»r lfl* '/ /«f"4*irJ   f^   "H /*   *' Xifi**

iM^^l     Jf>|      «0^,',     '^'4'#         r    -I   >***   **     *"   '*"'/   «'   ,H    '*


#     1%*  - f

>'*    Jf*;                    *


f I

768                             Appendix I

if not made "quite clear, are calculated to confuse the student
We will, therefore, explain each method as fully as need be.

ty«                                i. Duty.—This is the oldest method.    The boiler and engine

are considered as one machine, and the number of foot-pounds of
work done are stated, as obtained from i cwt. of coal^ the best
Welsh coal being generally used. The value of this quantity
having been found from time to time for the best engines (about
60 to 100 million foot-pounds), constituted a sort of standard for
the performance of others.

2.   Coal burnt per LH.P. per hour.—This, the later standard,
also including both boiler and engine, exists more or less at the
present time.   Its value was, at one period, as much as 4, but even
with single engines it was soon lowered to 3, in two-stage com-
pounds to 2\ or 2, and in triple-stage compounds to 1*5 or even
1*3.     The connection between standards i and 2 is shewn as
follows :—

Dutv  =         112 x 33,000 x 60

y        ibs. of coal per I. H. P. hr.
a relation easily proved by simple proportion.

3.  Steam used per I. H. P. per hour.—This is a standard of
comparatively recent introduction,   and  is  adopted   where   the
engine's performance is to be gauged apart from that  of the
boiler.     The feed   water   weight   is   measured,   and   priming
eliminated as much as possible.    The value may vary from  iz\
Ibs. in extremely good cases to 30 Ibs. with ordinary working; 24
being good for non-condensing engines.

4.  Efficiency of Boiler.—Just as the weight of steam used is
often taken to represent the efficiency of the engine, so the weight
of water evaporated is often called the efficiency of the boiler.
The evaporation per Ib. of coal may be taken at 10 Ibs. in good
cases.   A better statement of boiler efficiency may be made by

!il                         separately calculating the heat units given to the water and the

heat uiiit& obtained by the combustion of i Ib..of coal.    Then,

„ ..      ~ .             Heat units given to water

Boiler efficiency =   —==--------.2_____.._.

Heat units from coal

which ri&yliave a value of |o per cent.in a good boiler. (Seep. 1004.)



5. Efficiency of a Perfect Engine, that is, of an engine having a
reversible cycle. This is the highest efficiency to be obtained by
an engine working between given temperatures, and may therefore
be termed the ideal efficiency. It has already been shewn to be

where ra is the temperature of the furnace and r2 that of the
condenser. It includes, therefore, the efficiency of the boiler,
but, dealing only with the diagram horse power, does not take
account of mechanical imperfections. Supposing such efficiency
once obtained with given temperatures, the engine might yet be
improved by increasing the range of temperature, provided we
do not exceed the bounds of reason. Very often this efficiency
is taken for the engine only, with ^ as live steam temperature-;
but the steam being at a much lower temperature than the
furnace, a much lower ideal efficiency is possible. A very good
value by the first method is 77 per cent, and by the second, 32
per cent.

6. Thermal Efficiency of a Real Engine. — There are .two
methods of stating this, the relation of the work done to the heat
expended^ according to whether we include the boiler and engine,
or take the engine alone. Thus we have :

Thermal * efficiency (a) •«'**


Thermal  efficiency (b)

heat units in steam per I. H. P.

heat unjifs from coal per I. H. P.

Then a, good value for a might be 14 to 20 per cent., while
for b it might be 9 to 13 per cent Of course, if the thermal
efficiency is to be compared with the Carnot cycle, the latter
must be measured by similar temperatures. (See p. 883.)

7. Relative Efficiency.—This, as its name implies, is a com-
parison of the thermal efficiency of an engine with that of a
reversible cycle having the same range of temperature, and is the

*$« ^

I*     V^*


*^i   ^     |'.#f

tt    ^u* *%

f   t.

I*,             #1        010

^t** , **#                                    4*||  p

'i     |?S^|i*r|     rt|£     I*     ||»;

j,u *|<*>*t*.*Ai            >**   -*M*$* ,«u^

I II ^i   4| %             |

^#»   ^r*y*J   <-*t   n| *fw   pi   V t»9j+ tty*

*.     /1j P/ »•  »|«,//i«*,»f pf    ^«||           V^r^^J   jC^f«|>|,J4

; A         ,   i f  *    » i >
,**,**     , •?         '»       i j, | •

A   .     »i,   'i ;»^flf , ,^;f

« *      » t     . ^        f     > l»iS'f


* ft  f'^

Appendix I.


of expansion and compression between these temperatures are
•strict adiabatics.

The second and only* other method of obtaining reversibility
is to use a regenerator, as first practically attempted by Stirling, and
•afterwards by Ericsson in their respective air engines. Diagram
-A, Fig. 757, shews Stirling's cycle. The gas having a pressure



and volume corresponding to #, took in heat along the isothermal
-a b; rejected a portion during b c to the regenerator, at constant
volume; was compressed isothermally from c to d, during which
time heat was rejected to a refrigerator; and between d and a
again received the heat which was rejected from b to c. Although
«J> c and d a are substituted for adiabatics, the giving and receiving
of heat is strictly within the engine itself, the heat rejected, at b c
-being fully returned at d a, so the reversibility is unimpaired.
Diagram B illustrates the cycle of Ericsson, which is only
-different from that of Stirling in that the regenerator gives or
abstracts heat at constant pressure. As before, a may be con-
sidered the starting-point. * (Excepting as in Apfendix If.,
A 883.)

P. 613. Reversibility in the Steam Engine. — It
appears, therefore, that supply and rejection must be along
isothermals, and expansion and compression along adiabatics
(unless a regenerator be used). Remembering that isothermals
fbr saturated steam are horizontal straight lines, the reversible
cycle in Fig. 758 for the steam engine is easily understood.
Thus a b is the isothermal of reception, b c the adiabatic of
expansion, cdiht isothermal of rejection, and da the adiabatic

772                            Appendix L

of compression. Such an engine uses the same substance over
and over again, and the cycle is strictly reversible, for we have
complete expansion to <r, and complete compression to a. Its






efficiency therefore is measured like the cycle of Carnot.
substance is water at a, and steam and water at other times.


P. 623. Ratio of Expansion in a Single Cylinder.—
As a result of many recent practical experiments, it is found that
a ratio of expansion of between 6 and 9 is the highest limit
which can be adopted with economy, though a ratio of 4 will
yield practically as good results.

P. 624. Theoretical Diagram.-—In order to make the cal-
culations from the preliminary diagram agree with practical
diagrams, Prof. Unwin introduces a fractional coefficient in the
formulge, which he calls the diagram factor, and which is deduced
from experimental results of engines similar to that under con-
sideration. The value of the factor varies from '85 to '95 in
good engines.

P. 637. Exhaust Lap.—In the case of a D valve, an early
cut-off to steam will catfse an early cut-off to exhaust (see Zeuner's
diagram, p. 660). In such a case, if a later compression point be
advisable, it may be necessary njpt only to eliminate exhaust lap
entirely, but actually to give a small opening to exhaust when the
valve is at mid stroke. Such opening is then termed negative lap,
and would be shewn on the steam circle in Fig. 653.

Appendix L                                773

P. 650. An Isochronous Governor is a governor having
but one speed consistent with stability. A parabolic governor is
isochronous excepting for the influence of friction, its stable speed
occurring when the balls are in their lowest position. If this
speed be increased the equilibrium is neutral, because the height
of the cone does not change with the rise of the balls. All over-
sensitive governors hunt more or less, that is, they are apt to rise
too high and fall too low, even with the small changes of load
during a revolution, and the result is a condition of oscillation
which prevents a settled position corresponding to engine speed.
This extreme sensitiveness can be prevented by a spring (Fig.
645), but in a much better way by a dashpot (Fig. 261), for in
the latter case the air causes a constant though small resistance.

P. 700. Ignition Tubes.—Porcelain tubes have been used
for some time (1895), their life being about twelve months.
Iridio-platinum is also now adopted (1898) with great success.

Description of an Engine for Refrigerating Air.—
When work is done on a gas, without subtraction of heat, the
whole work is expended in raising the temperature of the gas;
and when a gas does work without addition of heat the whole
work is obtained by the abstraction of heat from the gas, causing
a decrease of temperature. In practice these results are only
partially effected, but the general changes obtained may be
understood by reference to Fig. 562, and P. 547, where
(theoretically) air at 60° F. is cooled to - 201° F.

The work derived from the expansion of steam in the engine
cylinder is employed to drive the piston of an air-compressing
cylinder, which draws air on the inner and compresses it on. the
outer stroke, up to 90 or loolbs by gauge. The temperature
rising to 200° or 300°, the cylinder is jacketed with cold circu-
lating water, to prevent damage to lubricants, and thus the air is
partially, cooled. Leaving this cylinder, it passes through the
pipes of a surface condenser, being there cooled to 60° or 70° by
cold water circulation round the tubes; and from the condenser
it enters an expansion cylinder covered with a non-conductor.
Here the air does work, helping the engine, and, cooling to



774                               Appendix I.

~ 10° or - 30°, is exhausted to the storage chamber, from
•which, after doing duty, it is redrawn by the engine to supply
the compression cylinder. By this arrangement there is less
heat to abstract than if air at 60° were used, and an important
•economy results. The compression and expansion cylinders
must lie apart from the steam cylinder and from each other, and
all the pistons are connected to one crank shaft, which has,
-usually, a heavy fly wheel. -The parts required are therefore, in
-order :—Steam cylinder, compression cylinder with water jacket,
condenser, expansion cylinder (non-conducting), and storage
chamber (non-conducting).

At first sight it would appear that heat abstraction was
entirely due to the use of condenser and water jacket, but this
is only a part of the truth. Heat given during adiabatic com-

x (diff of temp.)

diff. of temp.)
\n                           and,


i * |                            Again,   condenser   and   jacket   remove   heat   at   constant

"                       pressure, and the amount

ffjf]                                                      = Kpx.(diff.-of temp.)

Also, heat removed during adiabatic expansion

= Kv x (dirl of temp.)

Assuming, as in the theoretical example, that rise of temp.
r f/                       during compression = fall of teijip. during expansion, the various

;4                       heats may be represented as follows .—

N = normal heat (atmospheric temperature)

1.  After compression, heat in gas

= N + Kv   (latter given by engine)

2.  After condensation, heat in gas

= N 4- KV — Kp (heat now in condenser = Kp)

3.   After expansion, heat in gas

_XT i tr       tr   _ IT

— JL^I T JVv ~" -P^p       JXV

«= N - Kp   (heat taken by engine = Kv)

Appendix L


Let condenser heat go to boiler, and thence to engine; let all
expansion heat be abstracted by engine ; and let the total, together
with remaining heat in storage chamber, be used to compress the
gas; then

From storage chamber......Heat = N -. ELP

„    condenser and jacket    „    =       Kp
,,    expansion cylinder...    ,,    =       Kv

Total-  N + KV

which is the same as (i), or the cycle is complete, and the action*
(if without loss) would go on for ever.

Taking values for Kp and Kv and N as normal heat; the-
deficiency in N after condensation is represented by KP-KV =
53'2, and a further deficiency, shewn by Kv=i3o, occurs during;
expansion, the heat abstractions, being therefore 29 per cent, and
71 per cent, of the total abstractions, respectively. If the drop-
be less than the rise of temperature, the condenser will abstract,
a larger proportion, and if the condenser could take its heat at
constant volume, the expansion would do the whole work of heat
abstraction, for then the condenser abstraction would equal the-
compression supply. The condenser water goes partly to feed
the boiler, and partly to heat work-rooms, &c.

Fig. 759 is a section through the air cylinders, whose pistons,
are connected to the engine crank-shaft.    A, the   compression.


cylinder, receives air from the storage chamber, through E, which,
after compression escapes by F to the condenser. Thence it
re-appears, cooled to 60° or 70°, and entering the expansion,
cylinder D by passage G, the piston moves upward. The air
volume having shrunk by cooling, cylinder D is much smaller


f ; I

I -. '   .



i i

t ' ,

, f i



41 It
4 tl
9 fr|

4 ||
«> f|
, II

'.  t
* I

/*  4 * II ^     I  f |
. »   '.,1^ '   i   4   *    ill''

.If I

Ji .

; * i

'i- i i

'« k i




Pp. 3 and 42. Remelting of Cast Iron.—Experiments
made by Fairbairn in 1853, by melting Eglinton hot-blast pig up
to eighteen times, shewed that while at first there seemed to be
some improvement, there was a deterioration in the later meltings,
the iron becoming white and hard. Chemical examinations then
made by Snelus, and lately repeated by Mr. T. Turner cm* the original
pieces (lent by Prof. Unwin) indicate that the action is one of
oxidation, resembling that of the puddling furnace or Bessemer



of melting.






1 6

Combined Carbon.





	•    'I?

2 '00

P, 3. Moulding Sand.—Floor sand may have 6 parts by
weight of old sand to 2 of new sand and half a part of coal dust;
fadng sand, 6 of old sand to 4 of new sand and one of coal dust
Too much burnt sand, even if ground up, cakes when re-used, and
causes the metal to boil. Too much wetting is as bad as too
little, but the requisite consistency may be roughly tested by grasp-
ing a handful of the sand, which should just retain its shape when
the hand is again opened, A more scientific method is used by
Mr. Bagshaw,.who prepares bars of sand in moulds, by light
pressure, 12 inches long and i inch square, which he slowly pro-


Appendix II.


jects over a table, as in Fig. 760, till they break by the weight of
the over-hanging portions, the results being seen in the diagram.
Unless the sand be properly prepared by riddling, treading, and
wetting, no amount of venting can make a good casting.

P. 12. Four-part Box.—However intricate in form a
casting may have to be, the difficulties of moulding may always
be met by the introduction of sufficient boxes or of loose pieces.
Fig. 761 shews a method of moulding a three-legged pot or
*-skillet/ practised at the Carron ironworks for an almost un-
known period.. Four boxes are used, numbered i, 2, 3, and 4,
which give also the order of removal, there being partings at
a a, bby and cc. The handles are pegged loosely to the pattern
from the inside, and are afterwards removed in a downward
direction, the core being struck on a rough iron body.

As evidence of what may be done when given enough
moulding-boxes and loose pieces, there was exhibited at Paris in
1889, by the Socie'te' Cockerill, a single casting of 10 tons
weight, representing three marine cylinders, with standards, bed-
plate, feed- and air-pumps.

P. 14. Plaster Patterns.—The use of these is not difficult
to explain, their introduction here being due to their great
similarity to loam patterns. Fig. 762 will indicate the process,
it being desired to make a plaster pattern for a dome cover.
Firstly, a supporting mound A is made by the rotation of board
B over moist plaster-of-Paris, the vertical spindle hanging from
a wall bracket When dry, the mound is painted with shellac
varnish, and the thickness piece described by the board c. The
pattern D, thus formed, is afterwards removed, varnished, and
used exactly as a wooden pattern would be; and with care may
serve for a large number of impressions.

A further use of plaster occurs in moulding thin flat objects.
Imagine a flat cover, Fig. 762^, say for a sand-box. Firstly, a
wooden pattern, being impressed in sand as at A, and the parting
made, the upper box is filled with plaster. When the plaster sets,
the sand is removed, the plaster surface varnished; and the
bottom box similarly treated, as at B. Secondly, the boxes are
separated and the pattern removed. Thirdly, a new pair of boxes
is provided of the same dimensions, and, taking each box sepa-


Appendix II.

rately, a plaster cast is taken from each of the plaster blocks just
made. Now these carte, after drying, could not be fitted together,
for each is larger than its respective block by the thickness of the
casting ; but if each cast be separately impressed in sand, and the

two sand impressions fitted together, they will appear as at c, the
thickness space being left, into which the metal is run.

The casts having been varnished, will materially increase the
speed of moulding, for a pair of casts may. be divided between
two men, and the removal of the cast from the sand is more
expeditious than a thin pattern. The blocks may be retained for
future casts, but are of no use in moulding.

P. -27. Cubical Moulds in Loam.—Deeming it advisable
to illustrate flat mouldings in loam by at least one example, the
jet condenser, Fig. 763, has been chosen. The casting R consists
of a rectangular box containing a pump barrel, and having suitable
openings for exhaust, feed, and injection. It was formerly usual
to make complete patterns for such objects, but skeletons are now
largely adopted, the flat intermediate spaces being struck by loam
boards, In our case the wooden skeleton is seen in position in
the mould, being there shewn by the dense black portions, and
in describing its use we shall begin with the cope mould. An iron
plate A is placed on brick supports, and a coating of loam laid on,
which is then smoothed over with a plain board. The pattern
being set upon this bed, right side up, as at B, loam is filled in
round the side ; then, by using suitable striking-boards, as at 0
and c, the various flat surfaces are finished in facing loam, and a
pattern E embedded for the feed print. The skeleton is now
removed vertically, but the bottom strips FF, and the feed flange,

784                            Appendix IL

being loose, are afterwards removed horizontally. The top half
of the pattern is similarly treated, by embedding in loani as at
G, and the plates N and P are bolted together to facilitate lifting
and turning over. Here there are two loose flanges as at E, and
the runner and riser patterns, all of which have to be inserted.

The two halves of the main core are next made, the skeleton
being again used; but this time the loam boards are swept
outside it as at H, while the pump barrel is struck by the board j
for the lower half, and by a similar, but shallower, board for the
upper half. The flanges are supplied by pattern r. The two
main cores are, of course, made separately, and in order to remove
them the guiding strips for board j must be only temporarily
fixed, and the skeleton itself must also be split at the same
place. The cylinder core is struck on an iron barrel (seep. 14),
and cores are provided for the holes K K, Q, &c.

Lastly, the various parts of the mould are put together : first
the lower cope, then the lower core, the cylinder core, upper
core, and the upper cope, inserting the before-mentioned short
cores, and thus making ready for casting.

P. 31. Steam Cylinder in G-reensand.—Cylinders of
the smaller size cannot be moulded in loam like that on p. 21, so
a short account is here given of a greensand mould, aided by Pig.
764. The pattern is split horizontally through the centres of
steam chest and barrel, and is supplied with prints B and c to
secure withdrawal of the lower half, the sand pockets A, B, and c,
being afterwards filled by cores made in suitable boxes. A. core
u is also moulded for the steam chest, and the barrel core is struck
on a pipe M, as already described at p. 14; j showing the straw
rope, and u the loam covering. Examining the direction of
withdrawal of the upper half of the pattern, shewn by the arrow,
the steam and exhaust-pipe flanges are seen to be troublesome
parts. This is a case where some care in design would obviate
much after expense, for the flanges could easily be made to
draw if the pipes were equal to them in diameter; as shewn
however, they are loose on the pattern, and a third box is pro
vided for their subsequent withdrawal. One other method, at
F, requires ring cores between flange and cylinder body. The
port cores L L, pipe cores ft. and K, and stuffing-box cores D D>

Appendix II.


complete the mould, which, finally supplied with gates and runners
GG and risers R R, is ready for casting.              !

P. 34. Moulding Machines.—Besides the machine on p.
32, which has only one use, that of moulding wheel teeth, there

are many machines suited to general repetition work,, which is
thus done both more quickly and more accurately. The opera-
tions of hand moulding -are more or less simulated,,that is, the
pattern is first secured in position, and fastened to the box, the
sand is next rammed, usually with the box upside down, and







Appendix II


lastly the pattern is withdrawn upward or the box downward.
Woolnough and Dehne's machine, Fig. 765, provides for a raising
of the pattern, a turning over, and removal of box on a short
tramway. R R are the rails, clamped at a convenient height, x a
table on wheels, P a pattern plate, turning over when required,
and at other times clamped horizontally by screws A A. The
pattern halves x: and v are first screwed to this plate in mutual
correspondence, and the raising and lowering is performed by levers
M M, which act on pinions Q Q through shaft s, thereby moving
racks K K. The upper ends of the racks support sleeves L L,
which carry trunnions N NT, thus lifted or depressed as required.

The operations can now be understood. Assuming the box
on the top of the plate as at a} it is filled with sand and
rammed, the screws A. A and the cotter bolts preventing
rotation and lifting respectively, These screws are next released,
the plate raised, and the box turned through 180° into position £,
an intermediate raising being necessary, Lastly, the cotters are
withdrawn, the plate raised, and the box removed by the tramway.
This leaves the x half of pattern uppermost, and the previous
operations being repeated for it also, the boxes are bolted together
for casting.

To avoid the lost time due to raising and turning over, Mr. J.
Maclellan has devised a machine where P is rigid, and the box,
being always right side up, is filled with sand and lifted till it
meets the pattern. The ramming is then performed hydraulically,
by the raising of a second box of sand, which is pressed against
the first one, thus squeezing some of its contents through the
ribs and producing the necessary consistency. (Ste p. 969.)

P, 42. Whitworth Compressed Steel.—Seep. 790.


P. 44. "Woods.—Passing inward through a tree section, one
meets in order the bark, sapwood, heartwood, and pith. Trie
heartwood is best, and the sapwood should be ayoided if possible.


Appendix //.

The circular marks are called annular rings, while the radial ones
ate termed medullary rays, and the process of drying tends to
split the wood along the latter, as already shewn. To minimise
this fault the tree should be cut into balks before drying, and the
last should preferably be done naturally and gradually, over
some two or three years, during which time it is protected from
rain, but allowed free air-current Artificial drying or desiccation
produces more splits, or 'shakes' as they are called, and of the
latter, 'cap' shakes follow the rings, and 'star' shakes the rays,'
but the worst shakes are those that twist as they travel along the
log. Speaking generally, timber comes under one or other of two
great divisions—the pine wood, and the non-resinous or leaf wood,
Under the former we have all the softer woods, such as pines,
firs, and spruces; while the latter includes the hard woods, such as
oak, beech, elm. sycamore, ash, mahogany, &c.


P. 74. The Blast Furnace.—Before iron ore is smelted
it is often calcined or heated alone, either in open heaps or in
kilns, one heap of 2000 tons being kept hot for about three weeks.
Some ores, such as the Scotch, require no fuel for tHs purpose,
they themselves containing free carbon. An interesting method
is adopted in Styria, where the ore travels slowly down an inclined
kiln, the fire heat passing in the reverse direction. With the use
of the hot blast,-calcining is not so necessary.

The design of the blast furnace depends largely on the district,
that on p. 73 being from Cleveland, where poor ores and coke
fuel are the rule, and the dimensions large. In Scotland, where
rich ores and coal fuel are used, even a 6o-feet furnace is almost
too high, causing much trouble in keeping up the fuel at the
boshes. A large furnace may have an output of as much as 500
tons per week. The blast, air is heated to about 900° Fahr. by the
waste gases from the top of the furnace, and j:he nature of the
will -depend very, materially on the ore, that for Dowlais

Appendix TL


clay ironstone being (according to Bloxam) for every ton of iron

smelted—                , ~ . .     , ^                                  0

i Calcined Ore......        4$ c\vt

Grey Cast Iron   < Coal         ..........        50

( Limestone           ...        ...        17

( Calcined Ore
\ Hematite...
\Vhite. Cast Iron X Forge Slag

( Limestone




and one-third more air is supplied in the second case. Hematite
is a very pure red ore containing some 70% of iron, while the
above-mentioned ore will never have more than 50%,

JP. 74. Cast Iron (Effect of Elements).

Carbon exercises the greatest change in cast iron, and its
effects have been well explained.

Silicon, after carbon, is the most useful element. If present
up to 3!% it produces soft, strong, grey iron ; and if added to a
hard, whitish, and cheap iron, it will make it strong and grey. It
is now much used to improve poor irons, but should only be
present in small quantities in iron that is to be chilled.

Sulphur is prejudicial, causing blowholes : it should not
exceed '15 %.

Phosphorus is also harmful, for though giving fluidity, it pro-
duces brittlen ess if in excess of i%.

Manganese tends to dissolve the graphite and promote com-
bined carbon, and confers the property of chilling. If more than
i% it causes large crystals as in Spiegeleisen. Rapid solidifica-
tion also favours combined carbon (see Chilling, p. 34);

The following foundry irons are the mean of many good
specimens :



	i '42
	*5*   .


Appendix 1L

(Strong foundry iron.)

	Comb. C.


JP. So. Spiegeleisen owes its value mostly to the presence
of manganese, which varies from 3\ to 11£% at different times,
but less than 6% renders the material useless for steel making.
There is also about 5% of combined carbon, and the fracture
shews large crystals. Probably, too, the silicon present is of use,
for both manganese and silicon generally improve the quality of

P. 82. Whitworth Compressed SteeL—Apparently the
existence of blowholes in steel ingots is due to the fact that low-
carbon steel, when molten, is capable of occluding or holding in
solution certain gases. When solidification sets in, these gases
are liberated from the fluid only to be immediately imprisoned
by the solidifying steel, while slower cooling only increases
sponginess, for then rnore gases are given off. To avoid the loss
occasioned by the cutting away of the ingot head, and to improve
the rest of the metal, three principal methods are in vogue :

(1)  Chemical treatment {addition of silicon or manganese}.

(2)  Forging or Cogging {under steam hammer or hydraulic press}.

(3)  Compression           {when fliiid}.

Silicon and manganese diminish the gas released, and collect
such bubbles as are already formed, but they reduce ductility.

In the Whitworth process, a powerful hydraulic press is
applied to the molten steel when in the ingot mould, producing
a thoroughly sound material by the elimination of all bubbles.
The press is shewn in Fig. 766. The fixed head H rests on four
screwed columns DD, which are again supported oh the base-
casting A. The pressure water is admitted tinder ram B, which

Appendix II.


rises against the trolley c supporting the ingot mould R ; and the
load is resisted by the head F, being transmitted thereto by the
fixed plunger E. The head F is supported by rods PP attached
to the lifting plungers NN, and its upward movement is prevented
by the nuts GG, while through it there pass two rapid-pitched
screws MM, upon each of which is a wheel L gearing with the
nuts GG; and lastly there is a small hydraulic cylinder j, whose
piston moves a rack in gear with the wheel K, which again forms
a nut on the screw M. Supposing it be required to raise F to
admit the mould, the piston j is moved so as to turn K, and with
it M, thus releasing nuts GG, and moving them upward to a very
small amount. Then piston j is locked in the new position.
Next, the rams NN are raised, lifting F and also the screws MM
through wheel K, which is npw a fixed nut. MM thus revolving,
the nuts GG are moved upward at "the same rate as F. When the
proper height has been reached, the plunger j is moved back to
its original position, bringing nuts GG on to their seats to receive
the upward thrust.

The ingot mould consists of iron rings s s, in two concentric
sets, within which are placed blocks of firebrick TT, and a lining
of ganister. At each end, ring plates uu are fixed, to hold the
bricks in place, and the open mould is covered with loose pistons
Q Q, again protected by fireclay slabs. When compression occurs,,
the plunger enters the mould by the rising of the latter, and the
gases that escape through thie bricks pass upward or downward,
finally leaving by the holes in plates uu. A very high intensity
of pressure is absolutely necessary, less than 15 or 20 tons per
square inch being very doubtful policy. The press shewn can
exert a total pressure of 10,000 tons. (See third preface.)

P. 82. The Basic Steel process, known also'.as the
* Bessemer-Basic/ was introduced by Thomas and Gilchrist in
1886 for producing steel from phosphoric pig, which had pre-
viously proved useless for steel making. Its success is due to a
magnesia lining to the converter, obtained by crushing dolomite
or magnesium limestone that has been previously dried, mixing it
with tar, ramming it as a lining, and heating to 'coke' the tar.
When this 'basic' lining has been heated, 14 to 20 % of the
charge weight is thrown in as burnt lirne, after which the pig is


Appendix II.

added,, and the 'blow; proceeds till the iron is free of carbon ;
the ' after-blow' then takes place till the phosphorus is eliminated.
The converter is now- slightly tipped, some ferro-inapganese

added, the slag removed, and the-metal poured into ingots. The
percentage of silicon and sulphur should be very low, and there
should be from i to 2 % of manganese. If this be followed, the

Appendix II.


resulting product, basic steel, is the nearest  possible  thing to
metallic iron.     Finally, in the



Suitable for intermediate
pigs,  with,  say,  -2  to


Silicon supplies
heat for conversion

Phosphorus supplies

Heat supplied

Phosphorus must
be low

Silicon must
be low

•5% Phosphorus.

and the Basic process is chiefly of value on the Continent, where
phosphoric ores abound.

P. 83. Temper of Steel, or the proportion of carbon to
suit it to a particular purpose. The following table is the result
of actual analyses of Siemens' steel, extending over some three
or four years, and may be looked on as reliably representing
present-day practice. It shows the gradual tendency to decrease
carbon percentage, except in regard to steel used for cutting tools :



Forgings .


Percentage of

•3 to -4

•25 to -3


•15 to '18

Locomotive  wheel centres,

locomotive firebox girders,

marine stem frames, &c.
Crank   pins,   crank shafts,

connecting rods
Piston rods to stand wear

To weld well

Springs—laminated................................................    '4 to *6

Boilerplates—ordinary.............................................    '17 to '2

Boiler plates—"for welding   __•...................................    '15 to '17

Tool steel...............................................................  17

The percentage in Bessemer steel is somewhat lower, if the same
strength is to be retained. Wrought iron, also by analysis, con-
tains from "25 °/Q of carbon down to mere traces.

P. 84, Electro-deposited Copper.—Ever since the dis-
covery of electro-metallurgy it has been known that purer copper
could thus be obtained, but the metal proved insufficiently dense,
and very weak. These difficulties have been overcome by


Appendix II.

Mr. Elmore's process, which is principally used for the making
of pipes. An iron mandrel is placed in insulated bearings in
a solution of copper sulphate, and a number of unrefined copper
bars placed round it some distance off. The mandrel is con-
nected to the negative pole, and the bars to the positive pole
of a dynamo, and the copper is thus decomposed and deposited
on the mandrel at a rate of about "2 inch in 170 hours. At the
same time a piece of polished agate presses on the mandrel, and
travels, slowly from end to end backward and forward, so as to
cover the whole surface as the mandrel slowly revolves; the
copper is therefore being burnished as fast as it is deposited,
and is thereby made very dense and strong. When the pipe is
sufficiently thick, the mandrel is taken off and steamed, which
allows the pipe to expand so as to be easily removed. It is
stated that copper pipes thus made are 50% stronger than those
that are either brazed or solid-drawn, and have a superior
ductility; while further strength can be imparted by rolling.

Manganese Steel is obtained by adding ferro-manganese
to iron or low-carbon steel. The first attempt, with 2^ % Mn, at
Terre Noire about 1885, resulting in a brittle metal, the
experiments were abandoned; but later, Mr. Hadfield (1887),
by pushing the percentage higher, obtained complete success.
The results at various degrees are very curious, and are probably
explained by the presence of carbon, which is inevitable.


i\      : Produces no change if C be low.

3^ to 5 : Remarkably brittle cold, even with only '5% C, but not

so when hot.
5j to 6J : About the same.

6 \ to 7  : Strength   and  ductility increased ;    magnetic  quality

Very ductile.

If not very tough, can be improved by water quenching.
Too hard for filing. "   Strength equals crucible steel.

Entirely   lacking   in  strength,   no   matter   what   the


Practically non-magnetic.
Maximum strength.

9 to 10



Appendix II.


There is a rapid decrease in strength with any further increase
in manganese. It should be .noted that the carbon must be kept
down to i% in the 14% material, to do which the ferro-manganese
should have about 82% of Mn. No doubt if the carbon could be
sufficiently decreased, even 20 or 25% Mn could be added. The
ingots are from 28 to 30 cwts., and the ferro-manganese is
added in a molten state. Honeycombing is not bad, but the
centre of the ingot' pipes' considerably on account of the great
contraction of this steel.


Percentage                      (Water quenched).

of Mn.
10     : Ductility equal to mild steel, strength much greater.

13      : Still higher strength and ductility.

14      : Limit of manufacture: beyond this toughness decreases.

Finally, manganese steel is strong, ductile, and hard; free from
blowholes, and more fluid than cast steel, but pipes badly, and
requires good feeding gates to mould. Its hardness prevents
machining or fitting, and grinding only can be adopted. The
steel is suited to dredger pins and other articles subject to great
wear ; and is also useful for wheel tyres in conjunction with chilled
brake blocks, causing great grip. The following is an analysis :




and the strength and ductility are next shewn:

Percentage Mn.

2\ to 7| ••
2\ to 7j ••

'4       "

Breaking Stress sq. in.
Cast        •• 3^ tons
Forged    •• 25 tons

very brittle
3% elongation

38 to 50% elongation

(as against Basic cast steel: 22 to 27 tons: 26 to 30% elongation).

In the testing machine the material is semi-plastic from the
commencement, shewing permanent set with low loads.

Nickel Steel is a very valuable material, discovered about
1885, in the search for combined strength and ductility, which
carbon alone is unable to give to iron. But when it is said that


Appendix II.

•i% of nickel increases the cost of the steel by one-third, the
difficulties of introducing it commercially will be understood, so
that though the percentage Ni might rise -to 20 or 30 with
advantage, the practical limit is 3 or 3^, with which there is usually
'3 to '4% carbon. The following list, from actual specimens, will
shew that hardness is due to the presence of carbon :

Percentage Nickel. 2-05 2*62
 3*2 3*4
	Percentage Carbon.
 •19 •96
 *54 •16
 soft soft very hard medium hard soft slightly hard medium hard

		|    -      %&         "'"•
		^     \\
		jfc*             %\
		M                   %•
		1                  \^
All these specimens welded well, and could be bent cold.    As
regards strength and ductility the following are from actual tests :

Use, &c.
	Percentage Nickel.
	Breaking stress tons sq. in.
	Elastic limit tons sq. in.
	Elongation per cent.

Boilerplates   ...... Ship plates  .........
	22*3 26'7
	25 2O

No. 13 Ni Steel .... Gun tube .......
 4. 1 '6
	31*5 21 '2

Gun jacket .........
	2O M

No. 14 Ni Steel   |
 (for propeller shafts) J Gun hoops .......
 4.8 7
 1O* A
	27'5 2O'C

Wire ............... \
	*H*J /
	Ou 4
	zo 5

(easily drawn)    J

In other cases it has reached 40 tons breaking stress with 28
ftons elastic limit, and an elongation equal to mild steel. Compare
this with Forth Bridge steel at 30 tons breaking and 17 tons
•elastic, or the Eiffel Tower steel with 22 tons breaking and r6 tons
elastic. In all cases the nickel steel has the same ductility as
mild steel, but with 30% greater tenacity and 75% greater elastic

Appendix II.


strength. It resists shock, is very uniform in structure, flanges
well, and is less corrodible than mild steel. It is therefore suit-
able for shafts, propellers, ship plates, boiler plates, large guns,
and lastly, when surface-hardened, for armour plates. Evidently
the function of the nickel is to prevent the shortness caused by
carbon, while permitting and even assisting the latter to exercise
its strength-giving property; it is, however, unable to confer hard-
ness without the assistance of the carbon, but increases that
hardening capacity.

The Harvey Process is applied to armour plate to give it
such extreme surface-hardness as will resist the attack of shell.
The process is essentially one of part cementation, or the intro-
duction of carbon to a given depth, and has been practised in
England by laying the plate on a shallow fireclay box filled with
charcoal, luting with fireclay, and keeping at 2400° Fahr. for
several weeks. As the plates weighed 30 or 40 tons each, the
risk of breaking the boxes was very great. This objection is
removed at the Bethlehem Steel Works, U.S.A., where two plates
are hardened simultaneously, face to face, but with 8 ins. of
•charcoal dust between. They are then placed on supports within
a furnace, thickly luted with sand and fireclay, and gradually
heated up to 1700° Fahr., remaining at that temperature for 8 or
10 days, after which they are taken out and laid on supports in
an empty tank, so as to keep them apart, and allow water pipes
to pass between and around them. From these pipes a spray of ice-
cold water is directed on the plates for about an hour, and the final
cooling is done in an oil tank. The oxide on the plate surface
is afterwards removed by a pneumatic chipping-chisel (p. 949).

Armour plates are thus hardened to a depth of about if ins.,
and cannot be drilled or otherwise machined unless locally
softened by an annealing process. For this purpose an electric
current of large volume, from an alternating dynamo, is sent
through the plate at the required place, heating it to 1000° Fahr.;
.and the temperature is then let down gradually to a dull red,
which is tested by the burning of a pine stick in contact with the
plate. The electrical principle involved is exactly the same as
that in the Thomson process of welding (p. 329), the dynamo
providing a current of TOO amperes at 300 volts, which is changed


Appendix IT.

to 10,000 amperes at the plate, passing there through copper
terminals \ in, square, kept cool by water circulation. The
current is very gradually applied, and very slowly shut off, by
means of a rheostatic switch.

The value of nickel steel for armour plates is shewn by the
following figures, which prove a saving in weight of 43*8 % for
equivalent resistance, over ordinary steel plates:—

Kind of Plate.
	Relative penetration.
	Relative resistance.


Soft Steel         .........

Compound Steel and Iron

All Steel           .........

Nickel-Harvey  ...
	J 'OO


Chrome Steel is obtained by the addition of chromium to
steel having about -4% of carbon, and the resulting metal is
not only extremely hard, but is * self-hardening,5 that is, it only
needs to be cooled in a current of air after forging, to make it
suitable for metal-cutting tools or armour-piercing shells. For the
latter purpose the French Government use—


while other analyses give



Carbon °/0
	Chromium °/a
	Manganese °/0
	Silicon %

	not known
	not known



A 12 in. chrome-steel projectile has pierced a i6-in. armour
plate, while a chilled chrome-steel armour plate has received a

Appendix II.                            799

17  in. projectile,  fired with 8 cwt  of powder, the impression
being but if in. deep.    (See j>. 976.)                                 •       < ..

Tungsten Steel (called also Musket's Steel) is also a self-
hardening steel, similarly suitable for projectiles and cutting
tools, and having a composition by analysis of—

Carbon %            Tungsten °/0            Silicon °/0            Manganese °/0

1-36                       2-58                       -4.2                         -25

(Seep. 976.)

Sterro Metal is a brass to which have been added small
proportions of iron and tin. Its percentage composition is—

Copper......55 to 60

Zinc     ...        ...  34 to 44

Iron      ...        ...      2 to 4

Tin       ......      i to 2

It is both cheaper and stronger than gun-metal. Speaking
generally, the presence of iron reduces the tenacity, but in the
proportions shewn is of value. This metal has been used for
hydraulic pumps.

Delta Metal, though its proportions are unpublished, seems
to be simply Sterro metal in a forged or rolled condition. The
mechanical treatment thus received considerably increases its
strength, and it is much advocated for ship propellers. Though
the presence of iron causes a slight rust, the loss after six months'
immersion in an acid-impregnated water was but 1*2% as against
46% with wrought iron or steel.

Silicon Bronze is an alloy of copper and silicon, the latter
acting as a flux or reducing -agent, clarifying the copper and pre-
venting oxide scale. The resulting product is very strong, as
here shewn :

Elongation °/0

55   '

The second sample is deficient in toughness, and more than
5 % silicon causes great brittleness, Adding *r % of silicon to
melted copper produces clean castings by removing oxide, and a
little silicon to any brass or bronze is advisable. Silicon bronze

Percentage Copper.
	Percentage Silicon.
	Breaking stress, tons so. in.



Appendix II.

corrodes rather more than aluminium bronze, but is very close-
grained, and therefore suitable for resisting fluid pressure.

Aluminium is procured from its ores by one or other of two
processes. The older or Deville process has been much improved
as the Castner process, and is thus practised at Oldbury, near
Birmingham, being based on the displacement of aluminium from
its ores by metallic sodium. Caustic soda and iron carbide are
melted in furnaces at 1470° Fahr., and, thus being kept for some
i|- hours, the sodium distils over into iron condensers, and is
afterwards cast into blocks of 2 Ibs. each. Some 20 furnaces are
kept going at once, each producing 60 Ibs. of sodium per day,,
with an expenditure of 360 Ibs. caustic soda and 300 Ibs. iron

Next, alumina is prepared, by mixing the ground mineral
(bauxite) with soda ash, and heating it in a furnace till silicate
and soda aluminates are formed, after washing which, first with
water and then with hydrochloric acid, the hydrate of alumina
remains. This is mixed with common salt and charcoal into a
paste, and made into balls, which are thoroughly dried and
heated in earthen cylinders. While in this condition perfectly-
dry chlorine gas is passed over them, and aluminium bichloride
distils over. .»

Lastly, 80 Ibs. of the bichloride, 25 Ibs. of metallic sodium,
and 30 Ibs. of cryolite (another ore of aluminium) as a flux, are
heated together to 1830° Fahr.; and metallic aluminium to the
weight of 8 Ibs. is thereby produced, impure only to the extent
of a %.

In the Cowles process the ore is directly reduced in electric
furnaces, or rectangular fireclay pits kept hot by the current from
an enormous dynamo. Each pole within the furnace consists of a
bundle of five 3-inch carbons, having metallic caps or heads—of
iron if ferro-aluminium is required, and of copper if for aluminium
bronze. The furnace lining is made of lime and charcoal powder,,
the latter for localising the heat and saving the furnace materials.
The charge consists of ore, metal (copper or iron as desired), -and
charcoal; and the resulting alloys contain 15 to 17% of aluminium.
Aluminium has only a third the specific gravity of iron, and is
practically untarnishable. The addition of \ to i % to cast iron-

Appendix II.                            801

increases fluidity  and makes  casting possible with the whiter
irons.    (Seep. 1014.)

Aluminium Bronze, as a substitute for gold, has been
long known, but has only recently been advocated as an engineer's
metal. The best proportions are 5 to 11 % of aluminium, the
rest being copper; and the strengths are, with

5% Al.    24-5 tons sq. in., breaking.    40% elongation.
11% Al.    35*7 tons sq. in., breaking.    10% elongation.

If a small portion of silicon be added, the strength is increased
but the ductility diminished. The ro % alloy is much used for
bearings, gear wheels, propellers, &c. Shrinkage when casting is
very great, and good feeding gates are necessary.

Manganese Bronze has been mentioned at p. 85. The
ferro-manganese usually added is objectionable as introducing
iron, which decreases toughness and increases corrosion, so it is
better to use an alloy of manganese and copper. One of the best
and cheapest manganese bronzes has the following percentage

Copper    ...        ...    53

Zinc......        ...    42

Manganese         ...    3*75
Aluminium         ...    1*25

In view of the competition between the bronzes for propeller
construction, it may be noted that the relative cost for different
metals is given by the subjoined figures 4

Cast Iron
Steel ...
Delta Metal

Gun Metal
Manganese Bronze

Aluminium Bronze -

Phosphor Bronze


P. 98.    Steam Hammer Blow.—For the benefit of some
readers an amplification is here given of the matter on p. 98 -:—.,.

Velocity due to tup weight = ^/2^-H        (H = height of

.     ,    ^      , •    , -               Total pressure       PF
Acceleration due to steam =--------£—----- = -£..

mass              w



3 02                              Appendix II.

.-.   Velocity due to steam =  ^2/H =

+ V 2^H =

+ *

and Total Velocity

which value must be used instead of v on page 98.

s is more rigidly correct for mean total

Also P = —

pressure, though the addition is only slight. Thus, by calculation
a 5-cwt. hammer gives a blow of about 20 tons, half a ton of which
is caused by Ps, the pressure of the steam. (See p. 98.)

P. 124.. Stamping, — It is indeed remarkable how a difficult
forging may be overcome by the use of top and bottom dies, and
although the method ha.s its limits, it may be pushed to very

extreme cases, if we partly forge by hand and partly stamp under
a hammer. The metal should always be roughly forged or
welded to shape, however, before placing in the dies, so as to
dispose the fibre in the best direction for strength. Some seven
.examples are shewn: (i) The single-webbed crank, Fig. 12r,
Reeding no further explaaatiori; (2) the centre for screwing stock,

Appendix II.                            803

c, Fig. i2i#, which is first punched and roughed to shape, and
then placed in suitable dies; (3) the spanner A, Fig. 1210,
treated as in Fig. 767—that is, the jaw and shank are first scarfed
and welded as at A, then placed in dies ait B, arid finally punched
through at c before being removed; (4) the forked rod B, Fig.
121 a, similarly rough-forged, and punched in dies as at D, Fig. 767;
{5) the ring spanner E, Fig. 767, punched with a hexagonal punch;
{6) the hook F, Fig. 767, first roughly bent as at G, and then
stamped and punched as at H; and (7) the deep-eyed lever j,
Fig. 767, first prepared by two rough rings K K, and a scarfed
Tod L, then placed in the dies and drifted as shewn. Whenever
welding is done in the dies, the pieces must be raised to a very
good welding heat; see B and j, Fig. 767.

P. 125. Steelifying Iron.—This process is of the same
character as case-hardening, and is practised by making a powder
having i| oz. of prussiate of potash, \ oz. potassic nitrate, and
J oz. sugar of lead; placing it upon red-hot iron, and reheating
till the powder melts. Brightening a small portion of the iron,
the colour is watched for as in tempering, and the quenching
done in rain water. It is claimed that the hardening is very
thorough, and makes the material suitable for cutting tools.

P. 128. Hardening Steel.—The hardness produced in
•cooling steel depends very much on the rapidity with which the
heat is removed. Water is a good cooler, and is most used, but
much harder results are obtained by cooling in mercury, and the
hardest known by means of lead; the point of the tool, after
heating, being pushed into a block of cold lead.

Hydraulic Forging.—Advocates of hydraulic pressure for
heavy forging aver that steam hammers are done with, and that
no more heavy hammers will be ordered. Yet hydraulic forging
ivas used practically by Haswell in 1861, being proposed by
Charles Fox in 1847, and many big hammers have since been
built, proving that old prejudices die hard. At the same time
it is fully conceded that the hammer blow merely compresses the
•exterior of the forging, and never satisfactorily reaches the centre,
due evidently to the shortness of time occupied at each stroke.
The advantage possessed by hydraulic forging would belong also


Appendix 11.

to any method that substituted a steady pressure for a blow, only
that its rigidity makes water the most suitable medium. It is not,
therefore, introduced on account of its storage, qualities, and in
fact the use of an accumulator or any possibility of a blow is
disallowed at once. In most forging presses, then, the source of
power—steam—lies near the press, and the water is merely a
connection from there to the ram cylinder. The sole advantage
of the system is that time is given for the metal to flow right
through the forging thickness, and that this is no chimera, the
statements of most celebrated engineers admit no doubt of.

Apparently hydraulic forging was suggested by Whitworth's
compression of the fluid steel, and by the objection of his
neighbours to the hammer noise; but it is doubtful whether the
first practical press was due to Haswell, or to Gledhill, Whitworth's
manager. Fig. 768 is a plan of Has well's press, the steam piston


A being connected directly to the pumps; BB are non-return
" admission valves, and c c delivery valves, worked by powerful
levers from an auxiliary steam cylinder. The piston travels a
whole stroke in either direction alternately, valves cc being opened
or dosed as required, and the water Is exhausted through a fifth
valve, D. A smaller hydraulic ram placed above the main one
serves to lift the latter by means of links. Whitworth's press is
fed directly by steam-driven pumps, though the lifting rams are
worked from an hydraulic accumulator, and his apparatus is easily;

Appendix II.


understood from Fig. 769, A being the main cylinder and B B the
lifting rams.

Both methods have since been adopted satisfactorily, and two-
things have to be noted: (i) that immense rigidity of framing
is required on account of the heavy pressures, 2 or 3 tons per
square inch, and (2) that the difficulty of keeping valves and
packings water-tight causes some makers to dispense with the
former altogether. A very useful modern press, designed by the
late Mr. Tweddell, is shewn at T, plate xv, facing p. 318, where
the absence of pillars is a convenient arrangement

Cold drawing of metals has • long been practised in the
manufacture, of wire from more or less plastic materials, a hard

steel plate being drilled with a series of holes of gradually de-
creasing size, through which the material is passed in succession
from largest to least The principle is also applied to plates of
the same materials, which are stamped by a regular series of dies
until the required shape, often much removed from the original
condition, is attained Between every 'draw/ or nearly so, it is
usual to anneal the work, for such forced flow of material produces


Appendix II.

forittleness. The steel cylinders now used for storing compressed
gases are thus made in one piece; so also are boiler tubes and
•cartridge cases. The last-mentioned have become of large size
since the introduction of the quick-firing gun, and very heavy
presses are therefore employed, the operations at Woolwich in
•drawing such a case for a 6 in. quick-firing gun being shewn in
Fig. 770, as described by Sir William Anderson before the
Institution of Mechanical Engineers in 1897.

Metal-spinning is a method of moulding thin flexible metal
•sheets upon wood blocks fixed in a lathe chuck, by means of
a wooden tool or presser. In this manner knobs, teapots, and
many other domestic articles can be built from spun hemispheres
or saucers.


P. 152. Lathe Centres.—The American practice is to use
an angle of 60° for work up to 15 ins. diameter (7Jin. centres),




and 70° to 90° when above that diameter. Mr. W. H. Pretty,
Wh. Sc., of Bedford, writes that an endeavour there to use 75°
for small work (up to f cwt.), and 80° for larger work, met with
failure through changing ! of work, and a general standard of 80°
being first adopted, was afterwards altered to 60° to suit American
tools. For work above f cwt he drills the ends with brace and
bits, as in Fig. 771, by placing it on supports in line with lathe

Appendix II.


5, and giving a feed by advancing the poppet screw; the
irill being first used, and the countersink afterward.

jntreing   Machine.—This is a   simple but useful con-

:e for rapidly preparing work for the lathe.    The work is

in a concentric chuck A, Fig. 772, and a V support B ; the

being raised or lowered till the bar is level. The drill,
ig rapidly, is now advanced to the work by the lever, and
•untersink and plain hole both drilled at one time by means
: special drill-point shewn.

entre-grinding Wheel.—The common method of trueing
he centres is to soften them by heating, turn them up, and

v 77.4.


Appendix //".,

•then re-harden. Fig. 773 shews a handy emery wheel for trueing
up without preliminary softening. The shank A is fixed in the
slide-rest so as to let a vulcanite wheel E roll upon disc B fastened
to the catch plate, and the emery wheel c to just touch the centre
point, The knob D, loose on the spindle, is now taken hold
of to traverse wheel c, and the lathe mandrel is revolved, the
connecting band between spindles E and c being a long helical

JP. 157. Water-finishing Tool,—If a high and smooth
finish is to be given to iron or steel, a broad, sharp tool is used,
and plenty of water fed to its point while tooling. Quick speed
and large feed are also supplied, and the tool-points are shewn in
Fig. 774—A for a planer^ or shaper, and B for a lathe. The latter
form can be understood by remembering that the relative path
•of tool to work is that of a screw, while the tipping of the tool
at c permits its front to lie normally with the direction of travel,

P. 160. Face Lathe.—To give a clearer idea of this
machine, a general drawing is provided in Fig. 775. The
driving details have already been described for the break lathe,
and the slide-rest needs no further description. The only point
of difference lies in the bed, which, it will be seen, admits large
diameters having small axial width, It is thus a surfacing

J?. 260. Classification of Boring Machines.—A short
•classification will give a better understanding of the many types
of these machines. Thus, we may ha^e :

r, Boring in the lathe: with moving work,

2.  Special boring machine of lathe pattern.

3.  Horizontal boring machine: with fixed work.

4.  Vertical boring machine.

5.  Snout-boring machine, for blind holes.

Lathe boring has been described at p. 160; but as universal
tools are inadvisable, most machines being kept going with one
•class of work, it is better to construct a special machine (Class 2)
of c lathe-boring pattern, than to do much boring in the lathe

8 io

Appendix TL





itself. Such a machine is useful and expeditious, if the work be
not too heavy, and is illustrated in Fig. 776 by an example from
London Bros, of Johnstone, N.B. Some makers use two lifting
screws when doing heavy work, and others support the boring bar
on the saddle, as in Fig. 249, p. 237; but the main characteristics
remain, such as deep bed and short length, and the movement of
the work itself for feed.

Class 3, with fixed "work, is well described on p, 161.
„ The fourth class, the vertical machine, has already been men-
tioned, and its advantages, truth of surface for large diameters,
explained. The general design is shewn in Fig. 777, the con-
struction being similar to those of Class 3, where the work is fixed
and the tool fed along the bar hy an epicyclic train of wheels at
the upper end. Of course, the bar must be lifted vertically when
removing the work, but there is less risk of accident than with
the horizontal machine.

The snout-boring machine, Fig. 778, is made by Messrs. J.
Buckton & Co., for cases where a bar cannot be passed through
the work. "Very large diameters cannot well be done, but most
hydraulic cylinders can be conveniently bored. The driving is
by worm gearing, and the feed is given to the saddle on which
the work is bolted, lathe fashion. A facing head is also supplied.

JP. 268. The Slot-drilling Machine.—Fig. 779 shews one

FAJQ. 779.

of these useful machines, to Messrs. Buckton's  design.     It is
driven by speed cones A., from a countershaft  in order to make

Appendix IL


surface speed constant for various-sized drills. The horizontal
feed to saddle E is given by the gear on the left at B ; and the
driving gear consists of mitre wheels c and a long Marlborough
wheel D, thus permitting the tool to be set to any arranged depth.
The slide F, carrying the tool spindle, may be adjusted to give
the depth required, by means of spindle G, upon which is a small
pinion gearing into a rack on the back of the slide. The table
has the usual setting adjustments, and a hand feed is supplied
at H. The form of drill is shewn at j, which first makes a hole
of the proper depth, and is then traversed horizontally ; but if the
slot be very deep the work must be done in stages. (See p. 1019.)

jR. 775. Notes on Milling.—When work is being fed to a
milling cutter, the direction of motion of the work must be the
reverse of that of the cutting tooth. If this precaution be not
taken, the cutter will ride upon the work with great pressure, and
its teeth be broken. Thus in the left view, Fig. 181, the work

should be fed from left to right. Further, when using a helical
cutter, Fig. 181, the direction of helix must be such as to force
the cutter on and not off the spindle.

Heavy Milling Machines.—Milling is being more and
more adopted for repetition work. The mechanism of the Maxim
gun, for instance, is all but automatically turned out, by milling
machines of the pattern on Plate XII., intricate wavy sections
being cut by gangs of mills, or several cutters strung on one
4>indle. A not less interesting development is that of the heavy

812                             Appendix IL

milling machine intended to directly supplant planers, shapers
and slotters, even for ordinary unrepeated work. Over and over
again has it been proved that this can be done with economy,
despite renewal cost of mills, while the finish of the work is un-
doubtedly better. Fig. 780 is a vertical machine to do slotting-
machine work, and the horizontal machine Fig. 781 similarly
serves for planed work; the table movements giving feed in both
cases, and not cut. (Seep. 1020.)

Special Copying Machines.—The copying principle has
an extreme illustration in such apparatus as the copying lathe and
I!                                the profiling machine.   The former was the invention of Blanchard,

|                                an American, and was used by him to make such articles as shoe-

I] *                         makers' lasts, gun-stocks, &c.   It is still much adopted for ' turning'

*i                         the spokes and felloes of wooden wheels, and its principle will be

I i                       understood from Fig. 782.    There are two fixed headstocks, A

I /                       and B, and two corresponding poppet heads.    In B is placed a

cast-iron copy, say a spoke, and in A a rough piece of w6od. A
sliding carriage c carries a roller D and a fly-cutter E of equal
diameter, the latter being driven at high velocity by means of a
belt. The roller D is pressed against the copy by the pull of
weight F on the carriage, and the fly-cutter gouges out the wood
in imitation of the copy. The feed must also be given. This is
obtained by &very slow rotation of the mandrel B, which is com-
municated to the second mandrel through the idle wheel G, and
as the roller D is moved backward or forward, the same movement
occurs on the cutter E, so that the copy is accurately reproduced
at any section, whatever its shape. At the same time a slow
traverse is given to the carriage along the bed, thereby including
all sections of the work,

The profiling machine is similar in character, but is arranged
vertically, and is used for metal-cutting, its progenitor being
retained for woodwork. Referring to Fig. 783, A is' the copy, B
the work, c the milling cutter, and D a 'dummy' to traverse the
•copy, the carriage E being pulled leftward as before. An enlarged
view of the dummy at r>2 shews the cone shape which is required
to gradually increase the depth of cut, by refixuig at a higher
position after each traverse. The bed is long, and similar to thai                   f

of a planing machine, the feed being caused by a slow movement                   j

Appendix IL


of the  table,  as   in  milling  machines.     Flat   connecting-  and
coupling-rods are good examples of the work done.

fXa. 784.

Reversible Tools.—^The endeavours of early manufacturers
to increase the efficiency of reciprocating tools by making them


Appendix IL

cut on both strokes have not been highly appreciated by users.
Whitworth used a circular tool-box to his planing machines, and the
tool-point was automatically moved through 180° after each stroke
by cords and pulleys, much as the plate-planer tool is now moved
(see c, Fig. 285, p. 295). Lack of rigidity caused the abandon-
ment of this tool for good machines, however. Two recent double-
acting tools, by Messrs. J. Buckton & Co., are shewn in Fig. 784.
The tool A is suited to a slotting machine, and is automatically
reversed by rod B, on which are tappets. The main difficulty
here, and whenever one tool has two edges, is the difficulty of
sharpening symmetrically. The planing tool-box c is an improve-
ment in this respect, for both tools can easily be set at their
proper heights. The tools are fixed in slots made in discs, so
that when one tool is at work the other trails on the return, the
spring D keeping the acting tool to its cut. Both arrangements
are said to work very well in practice.

Facing Head.—Small articles such as pipe flanges are very
economically surfaced in a drilling machine by adopting a facing
head as in Fig. 785. It is fixed to the spindle by a coned shank
and collar as usual, and a radial feed is given to the tool-box D by
the star B and screw c, the former striking a fixed projection at
every revolution of the spindle.


P. 200. Turret-head Lathe with vertical Mandrel.—

The special tool shewn in Fig. 785^ is called by its designers,,
the Richards Machine Tool Co., a ' Universal Turning Machine/
but is apparently better described as above. It may be under-
stood by a reference to the description on p. 200. The mandrel
is supported on a footstep, and driven by worm gear, and there
are two vertical slides corresponding to a lathe saddle and slide
rest. The turret is shewn provided with tools for turning piston
rings from a cylindrical casting, for which purpose a turns the top,.
b roughs the thickness which c finishes, and d parts to correct,
width. For the last operation the tools are gradually fed through
the work by turning the hand wheel e. (Seej>. 978.)

P. 212.   Originating a  Surface   Plate. — Mr.  W.  BL



Pretty, Wh. Sc., finds it possible to save much time and labour
when originating surface plates or straight-edges, by giving the
workman a tabulated statement of the order of procedure, so
arranged as to systematically reduce the errors of manipulation.
The plates having been stamped with numbers, (i), (2), (3), in
conspicuous places, and the planing-tool marks eliminated with a
smooth file, he follows this order:

(a) Using (i) as a standard : bed (2) to (i); bed (3) to (i);

then bed (2) to (3), working equally on each. «
(V) Using (2) as a standard: bed (i) to (2); (3) is already
bedded to (2); then bed (3) to (r), working equally on

(f) Using (3) as a standard: bed (2) to (3); (i) is already
bedded to (3); then bed (r) to (2), working equally on

(d) Using (i) as a standard : bed (3) to (i); (2) is already
bedded to (i); then bed (2) to (3), working equally on

This cycle of operations is repeated until sufficient accuracy
is attained.   When straight-edges are being trued up they should
be  occasionally reversed  end   for end, so as to  eliminate all
possible errors.
V    P. 214. Screw-cutting. — There are no fewer than five





.Appendix II.

ways  of cutting a screw-thread upon a spindle, which may be
thus enumerated:

(i.) Cutting with ,$•&£& 0»/ dies as described at p. 192. This
principle is defective, for reasons there mentioned. The cor-
responding socket is screwed with taps.

(2.) Screwing in the lathe, described at pp. 147, 212, and 484.
This is the only method giving a perfect screw, except that next
mentioned, and is really equivalent to scale copying.

(3.) Copying is generally performed in a turret-head lathe as at
p;2ooj hut is also shewn at 47, Fig. 317, p. 349, for screwing
.stay tubes, The copy is often of a larger scale than the work.

(4.) By Chasing only.    On p, 212 the chaser is described a&

Appendix II.                              817

a finishing tool, when cutting screws in the lathe. It is, how-
ever, sometimes used both to start and finish the thread, being
held by the hand throughout the operations. It is easily seen
that such a method has precisely the same objections as stock and
dies, but even to a worse extent, for c drunken; threads, of varying
pitch, are often produced by inexperienced workmen. This is
because the tool has very short directing surface, even though that
surface may have the correct angle.

(5.) By Screwing Machine.—This method is the same in prin-
ciple as stock and dies, and is merely a quicker and more
automatic way of doing the same work. In Fig. 786 the general
form of the machine is that of a short lathe with fixed headstock A,
a clutch B for putting the mandrel in and out of gear quickly, and
a powerful driving gear. The work is held in a concentric chuck
c, and the screwing dies are in themselves a sort of concentric
chuck, operated by the lever E. The mandrel being revolved, the
cutting head D is brought over the work by the pinion and hand-
wheel F until the dies are in position to start: the lever E is then
pulled leftward by the advancing cutters, being partly helped by
the workman's hand at E. -The operation may be repeated until
the stud is of correct diameter, as indicated by the angle of the
lever E. With plenty of lubrication the thread may be cut during
one traverse, and when the dies are blunted they can be recut upon
a master tap placed in chuck c.

v Cutting Long-pitched Screws. — When the thread
angle becomes 45° or over, it may be questionable policy to cut it
in a lathe. Remembering that a screw is formed by an axial
traverse and a rotation, either motion may be called the cut, the
other being the feed; but the latter should always be preferably
the slower or smaller motion. Thus, in the machine for rifling
guns, the traverse being large and the rotation small, the tool is
propelled axially, the holder being a long bar with the tool
secreted in the end, in the manner known as the 'tiger's claw/
from the fact that it is sheathed on entering, and automatically
shot out on the outward or cutting stroke. During withdrawal the
tool bar is rotated by a rack and pinion, the amount of rotation
being fixed by the inclination of a bar along which the rack arm
travels. In such manner any long-pitched screw may be cut.


Appendix II.

But whatever machine may be adopted, there are some general
rules that apply to every case, though more particularly if the
angle be great or the pitch large, for then any deviation from rule
is more apparent. Imagine a screw A B, having, say, four threads
wrapped round its elementary cylinder. The combination of cut
and feed will cause the work to travel under the tool in the
direction K j, and the tool front must therefore be set normal to
this line. The section at H will not shew the real shape or width
of tool, but that at j, which is taken across the line CD. The
comparison of the two may be shewn by a diagram : thus, if ab
be the true pitch, over four threads, measured axially, and d E the
outside circumference, cd will be the pitch measured normal to
the threads. By setting out the threads on a £, and projecting
them on c d^ the true width of tool at thread top may be found,

y. 768.

as at Ci d?i. Again, by making d e equal to circumference at thread
bottom, we have fd as the normal pitch upon which the threads
are to be again projected from a &, giving f2 d2, the width of
tool at thread bottom. Finally, the curvature of tool point
will be that due to the ellipse obtained as section on line c D.
The same rules apply to broad traversing tools for plain lathe-
work, so far as curvature of tool and angle of path is concerned.

• jP. 226. Incorrect Taper-turning. — When cylindrical
work only is being turned, the correctness of the cylinder is not
in the least marred by the height at which the tool is set,
though the beauty of the surface may be very much affected.
But the turning of a tapered surface or cone requires special
"care in this respect, and it is of the highest importance that

Appendix II.


the height of the tool point should be exactly level with the
centre line of the work; for the section A B of a cone, Fig.
788, through the axis, is a triangle, for which a straight-line
feed would be suitable, but the section at CD, below the axis,
is a hyperbola, and a straight-line feed would simply produce
a reversed hyperbola instead of a true cone.

P. 249. Tooling Circular Arcs.—The methods of tooling
arcs of large radius may thus be classified :—

1.  Milling or slotting, with hand feed.

2.   Milling in a profiling machine having a curved copy.

3.  Special milling apparatus, p. 752, using a property of the


4.  Planing on a pivoted table controlled by a rod equal in

length to arc radius.

5.  Turning on a large face plate in a vertical lathe, whose

short mandrel is sunk in the ground.

Numbers i, 2, and 3 have already been described. No. 5 is
practised at Woolwich on racers or roller paths of large radius.
But as large arcs are not often required, it is better to fit up a
planing machine in the manner shewn at Fig. 789, where two


PlarvLrut ones

positions are given. Taking position L, let a line A B be drawn
immediately under the tool point, and let a stud B be fixed to the
table. Next, let a triangular frame A CD, one side of which is
the small table c D, be pivoted on stud B, and further pinned
at A; the hole in CD being slotted to permit the planing table to
travel, The work to be planed is now fastened to c D, and the
planing commences; then, by referring to position II., it will be
found that the curve c D, struck from A, will always lie under the
tool point whatever the table position.


Appendix II.


IR !

i I 1


P. 256. Turning Balls.—Brass and gun-metal balls are
much used for feed-pump- and safety-valves on account of
absence of sticking which they ensure. To turn them in the
lathe, a cup.chuck in hard wood—A, Fig. 790—is provided, into-

which the ball is fixed. This form of chuck permits the work
to be constantly changed in position; and, if this be carefully
done, it will be clear that any section will be a circle, which is

Appendix II.


the only requirement in a true sphere. After the ball is made
as perfect as possible with the usual turning tools, it is finished
with the tool B, made from steel tube, which is rocked to and
fro as the lathe revolves, the ball being often removed.

P. 274. Jigs.—These appliances may be constructed so as
to hold work for other operations besides drilling bolt holes, and
Fig, 791 shews two arrangements—A being a special chuck for
holding a dome cover while turning, and B a vice for supporting a
loose collar for boring.

Cutting Wheel Teeth,—The formation of the teeth of
wheels, either on paper or in the workshop, has been mentioned
at various places in this book as follows :—

Kind of Tooth.
	Cutting in Metal.
	Describing Teeth Curves.

	p. 60
	P- 31
	pp. 175 & 1 80
	pp. 510 & 517

	p. 60
	P- 31
	pp. 256 & 753
	P- 5*9

	P- 5*
	p. 10
	p. 274

A few more words on this very important subject will not be
out of place.

Spur-wheel Teeth.—A milling cutter is always used to remove
the interspaces, but the * blank' to be cut may be mounted in
one of various ways. On p. 180, the dividing heads are used for
support, but this method is only suitable for small diameters.
When wheels of large diameter form the regular work of a shop,
special machines are adopted; in many cases so constructed as to-
automatically rotate the wheel through the pitch arc after every
cut. Such a machine is shewn in Fig. 792, elaborate but effective,
where A is the cutter, B the work, c the dividing wheel, and D a.
sector for tilting the table to suit bevel gears.

Yet another method, shewn in Swasey's machine, Fig. 793, is
based on the fact that if a number of wheels be cut so as to each
gear with a given rack, they will all gear the one with the other.





$22                             Appendix IL

^The 'rack' is represented by the gang of cutters A, which simul-
taneously rotate and travel slowly in an axial direction. While
cutting takes place, however, the blank is also rotated on its own
axis, by means of change wheels, in such wise as to accurately
roll upon the (rack,5 with no slip whatever. The consequence is,
that the spaces cut out are not quite of the same shape as the
•cutter teeth, but are so widened that a real rack of the cutter
section would roll perfectly with the wheel thus cut. The axial
traverse of the cutters is given by the fixed cam c, which gradually


thickens from D to E, through about three-quarters of a turn, and
returns rapidly between E and D. Now the cutters are in halves,
and while one set F is cutting, and advancing axially, the other
set G, being out of the cut, is returning. The method appears
complicated, but in reality is very simple, and the wheels thus
cut will gear together most correctly, without backlash. By a
more recent method, a single emery wheel, of correct section,
rotates in one place while the wheel is really rolled along a line
parallel to the grinding axis; and thus cast wheels may be
trued up.

If a milling machine be not at hand, very good wheel-cutting
can be done in the lathe. The lathe centres support; a mandrel
which carries the blank, and the milling cutter is fixed on a

Appendix IL                            823

vertical spindle which revolves in a bracket set upon the slide-rest,
being there driven from a counter-shaft pulley overhead. The
whole apparatus is seen in Fig, 794, where A is the blank, B the
bracket, containing the cutter spindle c driven by the belt D, and
E a dividing plate which gives the correct pitch turn through the
worm wheel F.

Bevel-wheel Teeth.—It has already been explained at p. 753,.
that true bevel teeth can only be cut with a conical feed, and
that milling cutters are useless except for approximations, or
for roughing out before using an exact tool. (See Appendix V.+
p. 986.)

Worm-wheel Teeth.—These also  are cut-by many methods-

which more or less approach accuracy. The simplest way,,
though by no means a good one, is to rough out the spaces-
with a milling cutter, as on p. 58, and then to apply the worm,,
as there shewn, but using the file to trim down the thick portions-
of the teeth where marked by red ochre from the worm.

The second method is to use a hob (see p. 274). This tool is-
obtained by turning a steel worm of proper size and shape, cutting
out milling teeth upon it, backing these off at top and sides for
clearance angle, and finally hardening. Such a tool is highly
expensive to make, and there is considerable risk when hardening;,
and unless finished with an emery wheel is likely to be untrue.
Very few sizes can therefore be afforded, and wheels must be
kept to standard pitch. Again, supposing the hob provided,,
there are still two methods of applying it. The spaces are always-
roughed out with a cutter, in order to save the hob, and the
wheel being mounted freely on a stud, the hob is rotated while -


Appendix II.


being gradually brought into gear with the wheel. Usually the
wheel is not turned round automatically, and then the hob has to
•do the double duty of rotating the wheel and cutting the teeth.
The fault of this method is precisely that of stocks and dies in
•screw-cutting, for the worm thread has a different angle at the top



to what it has at the bottom. Better results are obtained by
using a machine like a short lathe, and mounting the wheel on
a vertical shaft passing through a fixed saddle, while the hob is
-driven between the lathe centres. The wheel shaft is then rotated
at the proper velocity regarding the worm, by means of change
"wheels.                          »

Appendix II.                                 825

:       The machine in Fig 795 was described by Mr. J. H> Gibson,'
*^ a paper read before the North-East Coast Institution in 1897.
.    is the driving shaft, connected to the cutter shaft B by change
Keels, arranged to give the actual relative motion between the
% <Orm and wheel.     The cutter D makes a purely circular rota-
v3^>n, which, however, traces a helical cut on the wheel blank N.
,   Ke shaft B is driven through nut E,- wheel F, and wheel G, the
riding on a feather key in the screw; but while cutting, all

pieces rotate solidly in the bearing H. Only once per
^volution of the wheel blank N, the lever p is pressed out by
cam Q, bringing rod R forward, as shewn by arrow, so that the
wheel j may catch R and cause wheel K to roll round G and F.
G has 42 teeth and F 40, these two wheels thus move relatively
one another, and the screw is shifted axially forward by the
pitch. The blank being fixed on the face plate, and the
p set just past the cam, the cutter is placed at L, and the
started. The result of the various motions is to make
^ series of light cuts or scratches on the blank, marking out the
l:r*terspaces ; but, when N has turned round once, the lever p
^oves, and, as previously explained, the cutter takes up a new-
Position on the helix of the imaginary worm. The cutting still
Proceeding automatically, the same grooves are simply cut a little
Deeper, and so the cycle of operations is repeated till the cutter
^merges at M, by which time every interspace will have been
Completely finished; for the cutter, representing the worm, will
have been presented in every one of its many positions, and
the wheel will have been truly 'rolled.' Finally the worm is
turned, with the cutter as template, and of a pitch equal to
screw BJ of which various sizes are kept.

•/ P. 277, Standard Fits. — Mr. Arthur G. Fuller has re-
cently made very careful experiments to determine the proper
zt/orking fit clearance, which should vary with the size of the
abject. As regards driving fits > the pin must be slightly larger
tfoan the hole, but the amount depends on the strength of drive
required. Force fits require a still larger pin, but the size would
again depend on whether the force applied were that of a lever,
screw, or hydraulic press. The last-mentioned fits Mr. Fuller
estimated from the average of all the experience he could collect,


Appendix If.

and the results of his investigations are set out in Fig. 796 for
working and force fits. Drive fits are to have from | to | the
largeness given for force fits, the former for light drives, and the
latter for heavy ones. The high and low gauges should be so







I *«M*t



*   •>*«j

MQ. 796.

NB.    DRIVING   FITS    OM£-HflL.F   FGRGE   f/TS,

made that the clearance (or largeness) should always be between
the maximum and minimum amounts shewn, and the method is
known as the 'limit' system. (For shrink fits see p. 842.)

f s


P. 286.   Caulking, if moderate, is not objectionablef but
split caulking is as bad as it can be, and should be
guarded against    It consists of splitting off and then turning in it
the joint a strip of iron about 7\. in. wide, and although it
stop a leak at the time, ultimately breaks out

P.  313.    Increased  Pressure  in   Rivetter   when

closing.-—The pressure of water in a rivetta:           1500 Ibt*

per square in. at first, is increased by about 50% at tbe

of closing the rivet, or to about 2250 Ibs* per             in.; the

Appendix II.


cause being the sudden arrest of the accumulator  weight and
consequent absorption of inertia.

^ P-.330. Electric Welding.—There are four systems at
present in use, the Thompson, Benardos, Zerener, and Voltex.
The first is useful for wires or bars, heat being caused by their
resistance, while the Benardos uses the arc, and is suitable for
repairs generally, but both require large installations. The
Zerener process avoids passing the current through the material,
the arc between two carbons placed in line being deflected upon
the work by means of electro-magnets. The Voltex system also
adopts two carbons; but they are set mutually at about 45°, and
magnets are unnecessary. Both the last-named systems are self-
contained, being carried about with ease, and it is claimed that
there is a saving in current of 30 % in the Voltex over the Benardos
process. Also that the double carbon apparatus does not harden
the parts so as to prevent machining, as in the other cases. The
Voltex requires a current of 80 volts, with 120 amperes for
welding, with 30 amperes for brazing, and with 5 amperes for
soldering. (Seep. 1.055.)

P. 332.    Setting for a Lancashire Boiler.—This type
of boiler not being complete without external flues, which also

SjeZ&rut of

. 79?.

	^             /

	s                                                         1^1
act as boiler supports, sections have been shewn in Fig.  797.
The direction of draught is indicated by arrows:  through the


Appendix IL


furnace tubes, underneath the boiler, and returning along the
side flues to the chimney. Where the brickwork touches the
boiler it should be as narrow as possible to avoid corrosion or
any unseen wasting, and should there be luted with fireclay
instead of mortar. The whole setting is lined with firebrick and
inclines slightly towards the front in order to drain.

P. 337. The Field Boiler.—This is principally interesting
on account of the Field tube, which has been much applied
where quick steaming has been required. The tube consists of
an outer or blind member, A, Fig. 798, having water inside and

furnace gas outside, and the inner tube B is supported by feathers
at the top, its funnel mouth entrapping the downward current,
thus causing the upward current to ascend by tube A, and pro-
moting the disengagement of steam. The boiler c shews the
tubes in position, hung from the firebox crown, D being a baffle
plate to retard and reflect the draught.

P. 339. Water-tube Boilers having been greatly favoured
for certain purposes during recent years, a more extended account
seems advisable. These boilers were tried in France as early as
1871, and have been used in the French Navy since 1874. In
the meantime the Babcock-Wilcox boiler was introduced in
various parts of the worldr The necessity for a boiler for the
English Navy suited to high pressures, combining light weight

Appendix II.                                 829

with safety, and having rapid steam-raising properties, created the
Thornycroft and Yarrow boilers, at least for the smaller boats;
after which the Belleville boiler (originally introduced for French
boats in 1879) was naturalised in England for the larger ships.
Meanwhile, the simultaneous development of the cylindrical or
' Scotch' marine boiler proceeded, and now it is a question which
is to hold permanent ground. The consensus of opinion seems
to be in favour of retaining the Scotch boiler for passenger and
cargo service, but especially for the latter, while the special
advantages of the water-tube boiler fit it for war vessels. On
land, the high efficiency of the Lancashire boiler and its steady
steam-supplying properties make it a difficult rival: neither is it
necessary to save space or weight on land. Taking now the
water-tube boiler alone, we may class its advantages :


1.  Will give higher pressure steam than cylindrical boilers.

2.  Repairs, though probably more frequent, are easily made.

3.  Steam can be raised more quickly from cold water, there

being little risk of deformation.

4.   Circulation is systematic, rising by the tubes, and returning

by 'downcomer/ Circulation, however, does'not in-
crease rate of evaporation, but the reverse: its function
is to keep steam moving and avoid dry plates.

5.  Lighter than other boilers for same power, because heating

surface is largely containing surface as well, whereas a
separate (and heavy) envelope is required for the latter
purpose in other boilers. Also heating surface is lighter
per square foot than in rnany other boilers.

6.  Less space for same power.

7.  Safer at high pressures.    This, however, is dependent on

good circulation and on a fairly small steam reservoir.

8.   Portability: can be carried about in sections and fitted up


The disadvantages of this boiler seem to be its doubtful
economy in regular use, and the fact that only smokeless coal
can be burned. Its greatest advantage is its favouring high


Appendix II.

I  *

pressures, which soon reach their limit in cylindrical boilers;
so, stated briefly, a less economical boiler is adopted to secure
a more economical engine. It is a mistake to suppose all water-
tube boilers fit for forcing:, the following classifies the principal
types regarding this quality:—

For Natural Draught only.



Lagrafel d'Allest.

De Naeger.


Capable of some Forcing.

\              Herreschof.

For Forced Draught.

Norman d.
Du Temple.

The four last-mentioned are most suitable for fast-speed war


vessels, such as ' destroyers,' which require increased combustion
on occasion.

The Babcock-Wilcox boiler has been already described, and
n|ay be taken as a type of natural-draught boiler. , The Ntdamse

Appendix II.


deserves mention on account of its use of an inclined Field tube,
the inner portion of which serves as downcomer; and there is only
one header, a double one, placed at the front In Fig. 799, A us
the rising tube and B the downcomer, the rising header c: entering
the stea.m receiver at a higher point than the falling header r>,
The tubes are easily removed from the front.

The Belleville boiler, now considerably adopted on  English
battleships, is shewn in Fig, 800.    A. is a non-con ducting casing,

limed with firebrick at B, and containing firebars cf ashpit i>, and
tubes E. The tubes are divided into eight or 'elements,'

of "which is a coil having its upper end connected to the
receiver F, and its lower end to the                       ci, which

Is a continuation of the downcomer H. The water,, auto-
matically controlled by a float-governed            valve,             the
receiver F at j, and is distributed along the lower portion, flowing

through pipes w and a, from the latter of which It enter*


Appendix II.

the element coils, rises, and passes into the receiver, beneath the
baffle plate which separates the hot steam from the colder feed.
The circulation is therefore caused by difference of density. K is
a mud collector, or blow-off chamber. It is usual to make steam
at a much higher pressure than is required, and let it pass through
a reducing valve on its way to the cylinder; this gives drier
steam by throttling, ensures a more perfect deposition of lime
and magnesia, gives a steadier pressure at the engine, and permits
the use of a smaller boiler for a given power. (See p. 919.)

In the Thorny croft boiler, Fig. 801, there is one steam receiver



A, but two mud drums B and c, and circulation is due to a rise in
the tubes D D and a fall in the downcomers F, outside the casing.
The feed enters at G, the water being delivered along the receiver
bottom, while two principal features are the entry of the water
tubes into the steam space of the receiver, and their tortuous*
form for the purpose of meeting the draught normally* H is a
baffle plate, and j the steam pipe.

The   Yarrow   and   Normand  boilers   are   similar  to   the
Thornycroft, but the first has straight tubes, while those of the
second are but slightly bent; in both cases they enter the
drum at the bottom.   The DM Temple is of the same type,   The

^ Lagrafal d'Allest, Orwlle, and De Naeger are similar to

Appendix II,                              #33

the  Babcock-Wilcox,  and the Herrtsdwf is a coil   boiler with
f>ump circulation.    (Saff. 918, 993,0;^ 1061.)

Plugs (see also p. 755).    Fig. 802 is a section of

one   of   these   as introduced  by the National  Boiler Insurance
Company.     The white-metal should be renewed annually.


JP. 367. Modulus of Resilience,— -The greatest clastic
stress is often called jtrocf stress. Then, from diagram 2, Fig, 326,
we may define resilience as ( half proof load x proof strain,' or


But A

-.  Resilience

and W

(A volume).

The quantity /2~E is termed the modulus oj resiliency and
measures resistance to impact per cubic inch of the bar.

61 .—Compare the resilience per unit volume of two bolts
A arid B, In which for ^ Its length A has a sectional area *8 of the
remaining ft : and B has for & its length a sectional area 'B of the
remaining $j (Hons. Mach. Canst. Exam. i8«9i).

In each case let/"« unit stress on smaller area
and 'Sj « nnlt stress on larger area.
Also let a « larger area, an<l *S a » smaller area,
Then, talcing "bolt A, total resilience


834                             Appendix II.

T28 f2 a     •
Resilience per unit volume = • — ^ ------- r v


Similarly, taking bolt B, total resilience

Jti»               2

^0*2 /2#                      /2

Resilience per unit volume =      ^ -- r- vol. = "4654^-

XL                                           iL

Resilience per unit vol. A __ 334 _
Resilience per unit vol. B      465              —

And bolt B is more economical for resisting shock, shewing why it
is better to turn down a bolt shank as at p. 402 to the diameter at
bottom of screw thread.

P. 369. Testing Machines. — The Drop-testing Machine
has always been favoured by railway companies for proving rails
and car axles. The apparatus is simple, consisting merely of a
large anvil on which are supports for the beam to be tested, a pair
of upright guides, and a falling weight that can be raised to any
definite height within them. The French railways use a * monkey '
(falling weight) of 440 Ibs. with a drop of u ft. 6 ins., the rail
supports being 3ft. yjins. span, and the anvil weighing 10 tons.
Messrs. Cammell use a weight of i ton, a fall of 20 or 30 ft., and
a rail span of 3 ft. Rigidity and inertia of anvil are of considerable
importance, but the chief difficulty is to attain constant rigidity.
The Pennsylvania railroad has therefore placed its anvil, weighing
17,500 Ibs., upon 12 stiff helical springs, each having two coils of
8" and 5^" diameter respectively, the outer one of i^" steel and
the inner of y£" steel, 9^" long uncompressed and 5^-" com-
pressed. These can support 80,000 Ibs., and exert a constant
resistance whatever the condition of ground. The monkey
weighs 1640 Ibs., its maximum fall being 43 feet, and the supports
are 3 ft. apart for axle-testing. It is specified that the axles shall
not deflect more than 5^-" with the first blow, delivered from a
height of 23!- ft, and shall stand five blows before fracture.

Testing for Hardntss* — Hardness may be defin ed as the resistance
to permanent deformation, often a property of great importance,
and a means of testing hardness (with accuracy is very much to
be desired All hardness tests depend upon making an indenta-
tion in the material by means of a harder substance, but the

Appendix II.


measurements are variously made. Mr. Thos. Turner loads a
diamond point till it just scratches, and the load measures the
hardness, while others have used points, knife edges or punches
of hard tool steel, and have severally measured volume, depth,
or length of the print. Prof. Unwin's method is more scientific,
for although he uses only a knife edge of tool steel, he has, by
plotting his experimental results, deduced the following formula :

Relative hardness = ~-

Where / = depth of indentation in inches.

p = pressure per inch width of knife edge.
n = 1*2.
His apparatus, Fig. 803, consists of a plunger A sliding in a


socket B, and pressing on the piece of excessively hard tool steel
c placed on the specimen D. The whole is loaded in a testing
machine, as shewn by arrows, and both load and depth of im-
pression measured, the former being gradually increased, and time
allowed at each step. In no case must the specimen be stretched.

(Measured by Prof. Unwin.)

Cast Steel        ............      554

Brass...............      233

Mild Steel       ...........      143

Aluminium (cast)        .........      103

Copper (annealed)       ...        ...        ...        62

Zinc (cast)        ...        ...        ...        ...       41

Lead (cast)       ............         4


Appendix //.

P. J7J-. Intensifying Compressor.—The differential hy-
draulic principle shewn in Fig. 727, p. 737, is a convenient means
of obtaining high pressure with small load, if little quantity of water
be desired, or a double-acting apparatus be permissible for con-
tinuing the stroke. It may equally well be applied to testing
machines for intensifying town's water, and is so shewn in Fig.
804, which represents Messrs. J. Stone & Co.'s apparatus for


testing pipe mains. The town's water is admitted through cock
A, and acts at B on the larger area, thus intensifying the water
in c, where the load is received on the annulus. The pressure
water then passes along pipe D to the main E being tested, and
the pressure is shewn at F. The cock A is so arranged that
town's water may be first admitted into c, D, and E, after which
the flow is diverted to B and intensification takes place, the ram
rising till its stroke is complete. If the test be not then finished,
cock G is closed, while c and B are placed in equilibrium by means
of cock A, and the ram is once more lowered. Again closing A,
the pressure is intensified instantaneously in c, and G being
opened, the test continues.

In a double-acting apparatus made by the same firm, a pair
of rams are used, one being up while the other is down, and the
cock, A, one to each ram, is opened and closed automatically by

Appendix IL


tappets, so that the testing is continued without attention until
stopped by hand.

P. j8jr. Test Specimens. — Professor Carpenter has
experimented on specimens of various lengths, 2", 4", 6", and 8",
from which he deduces that the ultimate strength (as shewn by
highest point of stress-strain curve) is independent of specimen
length, and that percentage extension (ductility) at maximum load
is pretty uniform. The ductility at rupture is, however, variable,
being inversely as length of specimen up to 8" long, and after-
wards constant. This indicates the need of a standard length
of not less than 8"; but if maximum load and extension only are
required, length is of no importance.

•P* 3#5-    Stress-Strain Diagrams.

Mechanical Hysteresis.—If tension experiments are made upon

rods or beams well within the elastic limit, by means of loads
gradually increased from zero to a maximum and then gradually
decreased to zero again, the ^ascending stress-strain curve will not
agree with the descending one. The first will follow the true
elastic line, but the second will slope more steeply and meet the
strain base at a point somewhat to the right of the origin. The
apparent permanent strain, called lag, will gradually disappear,
however. The area enclosed by the two curves has been called
mechanical hysteresis (from its similarity to magnetic hysteresis)


Appendix II.

and represents a loss which is probably due to heating. Fig. 805
shews experiments on wires 28 feet long, where hysteresis is very
clear. The iron wire had a diameter of "049", and the steel
wire '045". (Seep. 1071.)

Influence of Temperature, on Strength.—This may be separately
considered for low and for high temperatures.

At low temperatures Professor Rudeloff tried seven different
materials :

1.  Rivet iron.

2.  Rolled iron.

3.  Hammered iron.

4.  Acid open-hearth steel.

5.  Basic Bessemer steel.

6.  Spring steel.
Crucible steel.

And the temperatures for every material were three,  64°, -,

cj, 806.  Ijif/buuenjce/ of Jtemji^xJbuur/& on

and - 100° Fahr. Measurements were made of ultimate tensile
strength, yield-point strength, and percentage elongation in 3^ ins.,
the results appearing in Fig. 806, where each specimen is numbered.
Much, apparently, depends on the chemical composition of the
material, but generally the ultimate strength is raised rapidly at
first and slowly afterward, the yield point slowly at first and rapidly
afterward, while percentage elongation is generally decreased.
The material is therefore less capable of resisting shock at low
temperatures. Professor Dewar finds a , strength increase of
50 to 100% at - 295° Fahr.

At high temperatures metals decrease both in strength and
ductility. Copper alloys become .much weaker, but cast iron is
little affected. Probably Professor Unwia. has given most

Appendix II.


attention to strengths at high temperatures, and he states that
iron and steel usually gain slightly in strength up to 500°, but
after 600° decrease rapidly. For the copper alloys he deduces
the maximum strength:

/ = a - b (t\ - 6o)2

where the values of a and b are as follows :


	14. "8

Gun-metal (cast)  ...
	I 2X

Phosphor Bronze (cast)   ... Brass (cast)
	A ^ 0
 I 2*£
	•OOOO26 "OOOO24

Brass (rolled)        ...........
	A*  J 2 A' I

Delta Metal (rolled)         ......... Muntz Metal (rolled )       .........

Experiments are made by surrounding the specimen with a
box of oil heated by gas-jets, the measuring apparatus being
outside. (Seej>. 1074.)


	At 60° Fahr.
	At 400° Fahr.

Maximum stress per sq. in. Elastic limit per sq. in __        ... Extension in 10 ins.      ...        ...
	34'03 21-97

P. 390. Wohler's Law.—Unwinds formula, p. 390, being
empirical, cannot be proved from first principles, so its application
will be further shewn in the next problem.

Example 62.—Two steel bars, having a static breaking load of 30
tons per square inch, are stressed in tension, the one from 5 to 6 tons
and the other from i to 6 tons per square inch. Find the actual breaking
strengths to be assumed for the respective methods of loading,


Appendix II.


1°, tt

Now stress variation must be put in terms of maximum stress, and
this again in terms of/%, the new breaking stress, thus :

highest stress - lowest stress

highest stress
Case I. Ultimate static stress = 30 tons

S - 6~5/ - 1/'

D   —  —7-/2   -      /



27*25 tons square inch.
30 tons
• _ 5 *

2  —   6/2

.'./2 =

Solving the quadratic, /2 =
Case II. Ultimate static stress =


•'• A = A/2 + N/302 ~ (2 X |/2 X 30s)

Solving as before, j^ = 257 tons square inch.

The value of x in this formula may be 1-5 for iron and 2 for
steel in general, but varies considerably, the deductions from
Wohler's and Bauschinger's experiments being:


Bar Iron (Wohler)
Bar Iron (Bauschinger)
Bar Iron (ditto)
Plate Iron (ditto)


Axle Steel (Wohler)    ...
Axle Steel (Bauschinger)
Rail Steel (ditto)
^ ; Spring Steel (Wohler) ...
I Boiler Steel (Bauschinger)


P. jpj.    Average   Stresses.—The   following
figures are here given, all in tons and inches :




	Tensile Stress
		Extension per cent, at breaking.

Manganese Steel (14% Mn)

Nickel Steel (3% Ni for plates)    . . .

Nickel Steel (3.35% Ni for forging)

Silicon Bronze

Aluminium Bronze

Delta Metal (forged)        ......

Delta Metal (cast) ...        ......

Sterro Metal         .........

(Seep. 1075.)

Appendix II.                              841

P. 4.00. Thick Cylinders.—It appears from Lamp's formula
(p. 399) that in an originally unstrained cylinder, if the pressure
from inside be greater than f the tensile strength of the material,
no amount of thickness can prevent bursting. This difficulty can
be to some extent overcome by the principle of initial stressing,
p. 758. It has also been found that if the fluid pressure in cast-
iron cylinders be gradually increased beyond the calculated limit,
the internal diameter may be permanently stretched, and then /
has been known to reach 3 tons sq. in. with safety.

The formulae used in designing built-up guns, deduced from
Lame', is here given (see also Fig. 749).

/0 = internal pressure on firing

/x = pressure between A and B tubes on firing

/2 = pressure between B and c tubes on firing.

/0 = maximum hoop tension in cylinder A
^ = maximum hoop tension in cylinder B
/2 = maximum hoop tension in cylinder c.

r0 — internal radius of cylinder A
r-L = internal radius of cylinder B
r^ = internal radius of cylinder c
r% = external radius of cylinder c.

Then>    A - ~S> Co + A) +A          A =   aTa


To obtain these results, each outer tube must be smaller than
the next inner tube by an amount called shrinkage.

Q _ /Shrinkage between ) _ A ~A + 4 - *0
^2 ~ \         A and B         f ~         E~   " "      x

Shrinkage between I _     /2+A-(/0-A)

B and. c

I _


The formula for / may be extended to four or reduced to two

842                           Appendix II.

tubes, for it is seen how /0 follows from p-^ and so on, but for
shrinkage a further value is given for clearness, thus :

q       f Shrinkage between 1
b4= ]         c and D         J -

The radii are first assumed, varying approximately in geo-
metrical progression outwardly. A limit is next placed on the
hoop tensions, and p% found. From /2 we pass to p± and thence
PQ (the safe gas pressure) is deduced. Finally, the shrinkages are
calculated to give these pressures. A rough shrinkage rule is
given on p. 400.

v P. 400. Shrink Fits. — There are three methods of fitting                I

a   cylinder rigidly in  a  socket,  viz.   by driving, forcing,  and                 |
shrinking.    In all three, the cylinder must be larger than the

socket by an amount determined by experience.    The * largeness '                 |

for  driving  and force fits is given on p. 826; and for shrink                 }

fits, where the outer portion is heated before slipping over the                 t

cylinder, the following simple rules may be adopted : —                           t

S = '0025 x diameter of hole                                            /

if the parts are very thick and unresisting ; but                                            I

S = '0035 x diameter of hole                                            \

if thinner and more elastic. Care must be taken not to heat
higher than is absolutely necessary, and to prevent endlong sliding
by means of clamps. (Seep. 825.)

^ P. 402. Strength of Bolts. — Engineers have disagreed
considerably as to the stress which comes on fluid-tight covers,
though the problem is easily determined. First, suppose the
flange and seating be quite rigid, and no packing be placed
between the surfaces; also that the necessary tightness is obtained
by an initial ' sere wing-up ' stress in the bolt, as at A, Fig. 807.
If now the fluid stress be exerted, the flange wiH tend to take the
condition B, but as it is evident that (the flanges being rigid) the
surfaces cannot separate till the pressure exceeds the initial stress,
it follows that the bolt cannot stretch till such stress is exceeded ;
and the bolt stress must either be that due to screwing or to

Appendix IL

pressure, whichever be greater, but not both. As, however, an
excess of fluid pressure would cause the joint to leak, the initial
stress cannot be exceeded.

Secondly, suppose an elastic packing be placed between the
surfaces, and let us examine the problem by the elementary
apparatus at c, Fig. 807. Between two walls d and e a light
^block / is supported, by a thin wire spring b and a strong india-
rubber bar a] and let the walls be' separated so as to produce
a tensile stress, say, of 10 Ibs. in each spring. Further, imagine

that the wire spring is thus caused to extend J", while the rubber
bar stretches J". The stress-strain diagram is shewn at D, where
yy&=*iolbs., hj~Yi and/<£"=i"- Now let a force of 5 Ibs. be
put upon the block f so as to pull it leftwards, thus increasing
the stress in a and decreasing it in b. Shewing this on the
diagram, produce gk and/,6: make £/=5lbs., draw Im || k h^
and mn\\ Ij. The shaded portion shews the new diagrams^
indicating the stress in b as 9 Ibs., and that in a as 14 Ibs. If
now the force of 5 Ibs. be gradually increased, the stress in b will
decrease; and when a force of 50 Ibs. is reached as at hp^ the
spring is entirely freed from stress. The three cases are shewn by
the static diagrams E, F, and G.


Appendix II.

We next apply the elementary case to the model H. The
spring b becomes the packing q q, the rubber bar the bolt r; and
the two models c and H are therefore under similar conditions,
only that q q are in compression instead of tension. Applying
the same numbers, the screwing stress is 10 Ibs., felt equally on
r and ^, and shewn on dial /, and a force of 5 Ibs. being exerted
at u, there is a tensile stress of 14 Ibs. in r^ while the compressive
stress in q is reduced to 9 Ibs. Thus, any real conditions may
be ascertained by a diagram such as D, if the resiliences of bolt
and packing be known. If the packing be practically rigid, we
approach the case A, but the flanges always have a certain

In practice, the real difficulty is to find what stress the work-
man will cause in screwing up; hence the rules on p. 402 are
usually adopted, where the screwing stress is made a ratio of
the pressure, and the latter taken as the only guide in calculation.
Also the pressed area is measured to the inner edge of the bolts.

JP. 422. Shaft Couplings.—Mr. Archibald Sharp's coupling,
Fig. 808, is a combination of box and flange, and is an undoubted

X-A/o. or BOLTS

improvement on the latter as regards strength distribution. The
bolts receive shear stress along planes shewn by dotted lines, and
the twisting effort in A is transmitted through surfaces ab, be.
Similarly, B receives the twist through surfaces fb> be The
usual flange-coupling bolt has a shear stress over the whole cross
section, as t b c, but here ring c binds the outer halves of every
bolt, passing the strength ab to &£ %and the bolts are only

Appendix IL                             845

weakened at the half cross-section be.     Thus only about half
the usual bolt strength is required.

Equating the strengths of shaft and bolts, and neglecting the
small half cross-sectional bolt strength at <£,


The number of bolts may be X = 2 '5 J D

the nearest whole number being taken. If also the ring c be
made large enough to cover the bolt washers, it will have ample

P. 426. Deflection of Helical Springs. — In his difficult
and laborious investigations on the strength of square and
rectangular shafts, St Venant found that the greatest stress and
strain occurred at the middle of the (longest) side of the sections.
Now in formula for 0, p. 424, d is evidently the diameter of
greatest stress; therefore, for square wire,

wr                     , A       zs r             ,.



while for rectangular wire (see p. 421),

and A =          =            —    42.6

c*          ** *

P. 430-2: Moment of Resistance, Moment of Inertia
or Second Moment, and Centres of Gravity. — As area

has no mass, purists now object to the term * Moment of Inertia
of Area/ substituting the more reasonable ' Second Moment / and
* Centroid* is similarly adopted instead of 'Centre of Gravity of
Area.' Also we may define the nth moment of an area round a
given line, as that obtained by dividing the area into very small
pieces, and multiplying every piece by the nth power of its
distance from the given line ; or, algebraically,

nih moment = 2 (ahn)



Appendix II.

We thus have ist, 2nd, and 3rd moments of an area, each of
which has its use, and all easily found by a graphic construction.
Referring to Fig. 809, let the given area PPP be imagined to

L2    LI      L,

rotate round axis xx. Draw PPI II xx, and take any new line,
xtXi at any distance h, and I! xx. Project PXM and PN, and join
M N, giving a point QJ. Do this for several horizontal intercepts,
and obtain the shaded 1st moment area PCh In like manner
project Q^ and join I^N, giving the shaded 2nd moment area PQ2.
Similarly the 3rd moment area is obtained from area PQ2. Then,
calling the areas respectively A! , A2 , and A3 ,

ist  moment = Ax h   \

2nd moment = A9 h? > of the original area round x x,

3rd moment = A3 A3 )

and any higher-powered moment can be obtained by continued

Proof. — Let p P! be an element of area having width <?.
nih moment = S{PP   x e x (PN)*}




and LM

and ist moment = 2{ppx x e x (PN)J}
i x PN x e) — h x S(<? x PQi) = A^.



= -            and LLt


Appendix II



-,and 2nd moment = 2 {PPI x e x (PN)2}
(<? x // x PQX x PN) ...        ...    by substitution from (a)

(<? x #* x PQ2)         ......    by substitution from (£)

= ffi x 2 0 x PQ2) = A2/fr2

r moments may be similarly proved.

the construction to beam sections, we must first find
Imagine any given section ab (Fig. Bio).    Find its

-----^1   LINE    OF             \CU       LIMITING STRESS   j.



^b                          X

loment Ax, round xx say, using any height //. Dealing only
he right-hand half for simplicity, we have from the definition

A^ = A G                   Hence, G =  ——

gives the height of neutral axis z z by a much more simple
ccurate method than those on p. 432, especially if a plani-

is obtainable. We next require I, the 2nd moment, round
ting preferably the reference distance y to line of limiting

in constructing the curves. Treating the left side, to avoid
ion, every horizontal strip is referred to //, and its projection

to ^producing if necessary, till the original strip is crossed ;
hus the areas a± a\ are found, on opposite sides of the
L] centre line. Continuing the process on areas a± a\, the

848                              Appendix II.

2nd moment areas 02 a'2 are obtained on the same sides of the
vertical. Now the real value of I will be found by doubling our
results, for we have only used half of the section; hence,

I - 2(02 + a'2)j;2         and   Z - I = z^ + a'^y

or generally,

Moment of Resistance = /(2nd moment area)jy

if the reference distancey has been adopted. In cast-iron beams
jVt is always taken, which is less than yc (see p. 435), and the
resulting curves are slightly changed, but the construction ex-
plained must still be rigidly followed. It will be seen that this
method is superior to that at p. 430, though the latter is still left
in the text as sometimes convenient. (Seep. 1079.)

P. 441. Fixed Beams.—When beams have their ends
fixed as in Figs. 399 and 400, the Bm curve may be found
graphically, by supposing it the algebraic sum of two moment
curves, one caused by the ' action' of the free load, and the other
by the ' reaction' (a couple) in the wall itself. These opposing
curves must cover exactly equal areas: for, considering the
upper fibres of the beam say, the total effect of the load is to
shorten them; but, remembering that the total length of the beam
is unalterable on account of its fixedness, the reactionary couple
must entirely eliminate "the aforesaid compression. The sum of
the strains is therefore zero, and because Bm cc / oc b in uniform
sectioned beams, the sum of the Bm (average Bm x /) due to free
load must equal the sum of Bm due to wall couple.

Taking the special cases of Figs. 399 and 400, the first is
easily solved, but the second or uniformly loaded beam will be
further explained by Fig. 811. Now the Bm curve for a free
beam is a parabola, where b d, ~ W7-f-8; and the wall moment
is efgh) where eg = ^(bd\ to make areas A and B equal. The
final !Bm curve is shewn shaded at c. Lastly, to find contra-
flexure points, take any distance x between k and /.

t>     *         W/    ,.      W7   /W        *   r   x

Bm at x =----D = — - I— x.- -W • -

12              r»2    \2         /        2

Appendix IL                              849

.v be increased till it equals kL   Then BM1 = zero, or

from which, solving the quadratic,

x or $/ = "2i il.

Let a fixed beam A B be loaded in aay general or unsymme-
iriral manner as in Fig. Si 2, The curve of B,n for free load is
fcnmd at abc. The reactionary moment curve defg is next to
ins obtained. The two areas rz/'rant! defg must now not only

t»t* i.'(}ual, but their rentroids nuist lie on a common vertical.
If then x shew the value of the free area ct^as also the position
wf its rcntroid, v and "/, tiuwt be placed at -|/ and of such value
«* to             x in the manner of parallel forces, -where y 4- z « x.

'fhcn area /// - ¥ and        dgf^t^ from whieh the moments ^^

!/«» be ciedwred.     Lastly the resulting moment curve Bf
w «t)taiiii!(I by sujperposition.

VIIL, Fig, 401, p. 443, k *hevrn in ciowbl^ at the top

of Fig, 81$ next

Bending Moments for Continuoms Girdera
—The                        may be farther              to these

Appendix: II.

Four examples are given in Fig. 813, where for two spans we
merely have a duplicate of Fig. 401, p. 443. In all cases the
parabolas W / -v- 8 are first drawn for free beams, and these are
next opposed by a curve of reactionary moments, consisting of
straight lines shewing zero at the extreme ends. In the figure the
actual moments are drawn and stated, as found from Clapeyron's
formula, the shaded curve giving the final Bm; but the second or

opposing curve may be very closely obtained in a graphic manner.
Divide every span in thirds, drawing vertical lines, and on every
vertical except the outside ones put a mark at f of W/~ 8. Then
pass the second curve through these points so as to give the best
average results, which is easiest done by string, pins, and weights,
as at A. This method, which is due to. Claxton Fidler, is clearly
an extension of the principles already explained, and has the
advantage of being applicable to any mode of fairly uniform load-
ing, however much it may vary from span to span. (See p. 953.)

Appendix II.


Graphical Calculus.—Let a curve be drawn with given
ordinates and abscissae, and be called a primitive curve. A
second curve may be constructed to the same base, which will
shew the gradually increasing value of the area under the first
curve, as we travel say from left to right; and this second curve
we shall call the sum curve. It represents the integration or
summation of the first curve. A third curve can next be drawn
shewing the rate of rise or fall of the primitive curve, and this we
shall term the rate- or slope-curve, being simply a graphic differ-
entiation of the first curve.

Let the primitive curve A B, Fig. 814, rise from j^ tojy2 while

the abscissae change from.% to #2.    Then y increases jy2 —y\ units
of y for ^2 - #! units of ^?, or

Rate of growth       | _ y% - yl = d _

of j, for one unit of x }


This is the mean rate of growth between a and e, and may be
assumed to occur aty; the midway point. The instantaneous rate
may be found by supposing b to gradually become indefinitely
small, when the line a e will finally assume a position tangential to
curve A B, at the point considered. Hence the rate of growth of
a curve is always shewn by the trigonometrical tangent, or the
slope, of a line drawn tangential to the curve.

To draw the Sum Curve, that is, to find area A B c D under

852                             Appendix II.

any primitive curve, Fig. 815 : erect several vertical lines,
unequally spaced, there being more where the curve varies greatly,
also further mid verticals n', 22', &c. Take any pole o, at
a known distance p from a vertical XY : project points i', 2',
3', &c., horizontally to XY, and join each projection to o.
Next draw A# II 01", a6\\o2", bc\\Q$", &c., till the whole
sum curve A to k be formed, whose ordinate at any point will
shew area under primitive curve up to that point, measuring from
A ; while DJ will represent the whole area A B c D.

To Draw the Slope Curve, the reverse process must be per-
formed, and thus the primitive curve in Fig. 815 will shew by its
ordinates the slope of the sum curve. For example, suppose we
want the slope between c d, we draw 04" || c d, and produce 4"
to 4', giving a point in what will be the slope curve.

For proof draw cc± horizontally.    Then

slope vied = dc± = ill = vertical of^lope curve
cc^       p                     p

If we make /i = i, the steepness of any primitive curve will be
measured by the slope curve ordinate. Again, by cross multipli-

But cc± x 44' is the area between two verticals of primitive curve.
Hence if/=i, dc^Qi rise of sum curve will shew increase of
primitive curve area, and ordinate of sum curve at any point will
shew area of primitive curve to that point. Also the value/ may
be so arranged that the new curves may be read off directly to any
chosen scale.

Bending Moment and Vertical Shear by Graphic
Summation. — The shear on any beam section is due to total
load on one side of section less the support reaction (if any) on
that side. The beam A c, Fig. 816, is stressed by a load whose
intensity per foot run is shewn by curve ABC, drawn to a scale of
100 tons per inch. Base scale being i foot per inch or ^, let
, f in. shew 100 tons on shear scale. Supposing an increase of 100
tons on the shear curve, over a base of i foot; d = 100 tons, £ =
- in,, and y (load) = 100 tons = i in.

But   L = £
.>'      ^

Appendix IL

. ._»_



and the polar distance is fixed. Taking base D M, summate ABC
from pole o, obtaining D E F M, shewing total load on the left of
any vertical. Next make D H = M F, and with H as pole summate
D E F to base D j, obtaining D G. A vertical through G will pass

through the centre-of-gravity of the load. Set off a c^ A c, ag^ A G,
and draws£- K horizontally. Then D L and F K are right and left
reactions respectively. The deduction of area DLKM from
D E F M gives the two triangular areas, whose verticals indicate
shear, or total load on left of section less left reaction.


Appendix Ji

From a study of pp. 4.38 to 442 it will be seen that the Bm
curve is simply the surn of the shear curve, the point of origin of
the former being where Bm = o, viz., at the supports of a girder,
and the outer end of a cantilever. E^ will be a maximum where
shear changes sign, and will decrease with minus shear. In Fig.
816 let Bm increase by 1000 ton-feet over a base of i foot, caused
by a shear of 1000 tons intensity over i foot base; and let a Bm
scale of -Jin. = 1000 ton-feet be proposed. Then ^'= 1000 t.f. =
-j in., b' = i in., y' — 1000 tons — •§• in. x TO, and



= 5m-

The Bm curve can now be drawn, on base N P, by sunimatmg
D L E K F from pole or     Commencing at P, the curve rises to Q

£Ot/v/»/vc'V /A/

the maximum, and then decreases from Q to N, the shear being
minus. These constructions are especially useful for ships. A
curve of weights being drawn, is opposed by a curve of buoyancy
or the weight of water displaced at every section, and the net
result is the load curve, from which shear and Btt may be deduced
as before. Two extreme cases are taken, one with wave crest
amidships, causing 'hogging' strains; and the other with crests
near the ship ends, causing * sagging' strains. Lastly the sections
are treated as built up beams. Fig. 817 is an actual example.

Appendix II.

P. 446. Culmann's Diagram.—To prove this construc-
tion refer to Fig. 818. Taking any section as B, we shall shew
that Bm = LU x H. Reaction x = G j, triangles K L M and o j G are

fV Q5     Q*    «*

U V   I   +=



similar, as also are N p M and o Q G ; and so on with c s P, D T s,
and E u T.

9.9 = p]^       and QGX/ = PMXH.. ...... = moment of 6.

H         p

Similarly    5 x/j = SP x H......... = moment of 5.

and    4 x/>2 = TS x H......... = moment of 4.

and    3X/3 = UTXH......... = moment of 3.

M L        _ ^ „,.,.„ _ M L x H = moment of x.

Again,   -    «

6       '


.'. Resultant moment = M4 - (M/6+ M/5 + M/4

= L u x H.

It is wise to call L u distance and H force, thus keeping
space and force diagrams quite separate. Shear is easily proved
from previous statements. (Seep. 1085.)

P. 4jo. Deflection of Beams by Graphic Summation.
—-Let a beam A v, Fig. 819, be bent to the curve A B c Y, by means
of a load of any kind, the beam section being of any shape, but
here considered uniform at all sections. Imagine a small portion
A B, in which the external fibres DD will be extended or compressed
by the amounts d and d± respectively; and if b = strain,


Appendix II.

Let lines a and b be tangents to the curve at A and B respec-
tively.    Then 0 = a, and tan a = a, as a is small.    Also § -•

I>=j;0=J'tt==:j/ tana

and 7 r>'= ty tana = Ejy (slope accumulated from A to B)

Let Q H j be a Bm curve for the bad, which we can transform
into a limiting-stress curve, for/= Bm^-Z.    Then

/D = area M = (slope accumulated from A to B) Ey

also,  area NT == (slope accumulated from B to c) Ey

.-,  areas M-F-N- = (slope accumulated from A toe) Ey.

Hence,     sum of stress    ,11 total slope at )' *
curve to any point )       (   that point    J • •?

Appendix II.                             857

And, slope ordinate = (stress sum-curve ordinate) —


Thus the slope curve s may be drawn by summating the stress
curve, and then dividing by Ey.

Again, if A B be made very small, the angle a = VA B, and tan a
= average slope between A B. Then,

Deflection at B = D tana = D (average slope A to B).
Similarly  the   angle  (a + /3) = inclination  of  B c   to   A v,   and
tan (a + /3) = average slope between B c.    Then,

Total deflection at c = D tan a -I- F tan (a -h/3)
= D (slope A B) -f F (slope B c) = ef+gh

And, sum of slope curve 1 __ f Total deflection
to any point           J """ \   at that point.

The problem is therefore completed by drawing the deflection
curve A as the summation of slope curve area, and the general
formula is deduced for a beam of uniform section :

Deflection - *<«=) _ S(SBJ
ZEy           El

To prove that the maximum deflection of a cantilever with
concentrated load = W/3/3El (p.451): the Bm curve is triangular,
whose sum = W/2/2, and this is the maximum slope. But the
slope curve is a parabola, whose area is therefore |- W/2/2x/
viz., W /3/3, which is the second summation of the Bm; and by
above general formula, deflection = W /3/3 E I.

Some care is required in fixing scales, but previous explana-
tions may be consulted. When beam section varies, Z will alter
also, and the stress curve must be found : then continue as before.
Note that load^ shear, Bm , slope^ and deflection curves are a con-
tinuous series, where each is the sum or integral of the preceding

P. 458. Pillars and Struts.—In the paper cited on
p. 458, Prof. Fidler assigns various reasons why pillar strength
cannot be shewn practically by Euler's formula, such as an in-
constant E, even in the same strut, and initial curvature in line of
thrust, the latter altering W considerably. In Fig. 820 the crosses
shew Christie's experiments on T bars, and the small circles


Appendix II.

Hodgkinson's results for hollow round W. I. columns, co-ordinates
being/and r, as on p. 459. As the plottings scarcely approach
the Euler curve, the problem is to find what curve will suit.
Gordon's curve G with a = 16 and b = -^y^ evidently strikes an
average, which he intended it should. Fidler objects to this
treatment, holding that the curve should be made to fit the lower
or worst results, and he has therefore devised a formula, too
complicated for regular use, and shewn by curve F F. Apparently,


££   EuJL&r Cu.rv& .•   £*26toooooo

FF    fAJcULer   CuLTUCf   fx£t&aU to Lou/er

G G     CrOrdLoTLf C&tJ'UC,          OL,«    16         1} s  jtooo

9    Cordon, Curve,         cc,-   fj      b ~ lofco

Fig. 820.

however, a judicious alteration of Gordon's constants should
approach Fidler's curve, for when a = 17 and b = ^5^ line gg
is drawn; but if a = 14 and b = T^9 the lower and safer line
hh is obtained. Another way is to take /some 80 or 85% of
Gordon, which would meet the experience of engineers.

P. 461. Combined Torsion and Bending.— We shall
here shew how the two stresses ft and j£ are combined into one
equivalent stress fc. If a pair of shear stresses Fx F1? Fig. 821, act
on an imaginary solid A BCD, within a structure, they cannot

Appendix II.


exist alone, for they cause a turning eifect. They must, therefore,
be balanced by a second pair of stresses F2F2, where Fxx AB =
F2 x A D, or (/! A D) A B = (f2 A B) A D, and^ = /2.

Let us next imagine a second block A B c D, Fig. 822, acted
on by a shear stress /s and a direct stress ft. These may be
balanced by the dotted stresses, but we shall consider them
resisted by f& and f0 on plane c B. Now if the value of d be
properly chosen, f0 may be entirely eliminated, and the forces

on B c be solely, direct, as /e.     Assuming  this  condition,   and
resolving   ft c B   on c A :

(/eCB)sin 0 = (/tAB)-(/s CA)
Dividing by c B :       /*. sin 6 = ft sin 0 - f$ cos 6

and (/e -ft) sin 6 =  -/s cos 6.....................    (i)

Resolving (/i c B) on A B :

(ft c B) cos B =  - /s A B
Dividing by c B :   y^ cos 0 =  — j^sinO   ............    (2)

Multiplying (i) and (2) together:   /e2 -/e/t = /s2    '

and solving the quadratic :      /e « J (/t 4- V/t2 + 4/s2)
Inserting the values ofj^ andj^ on p. 461:




Appendix II.

which is the so-called equivalent twisting moment, although /e is
a direct stress and not a shear. The result is only true for round
shafts, but can be used for other sections by adopting proper
values of Z.

I :


P. 4^4-8.   Velocity  and Energy  Curves.     Let any

velocity curve, A, Fig. 823, be plotted to a base of equal times;
then  the  acceleration / will be   shewn   by  the   slope  curve

(see p. 852) of v, and v will be the sum curve of f, for velocity
growth v =//. Hence acceleration is rate of velocity change
regarding time. Again, at B, if an acceleration curve/be plotted
to equal distances, fd^^v* (for v* = 2/#). Therefore curve
^ #2 is the sum of f regarding ^, and the double of these
ordinates is #2, from which v may be found. Conversely, v being
known,/may be obtained directly, as on p. 492.

A curve of force at c will give an energy curve by summation,
for energy = p d. Many applications occur: thus, if / be an
indicator card, E will shew foot pounds (W^) given to piston, and

;                                               Appendix II.                             861

*                if E shew energy as delivered from a rifle bullet (W#2/2 g) while

A,               penetrating a target to depth 4 the pressure exerted at any point

will be shewn by the/ ordinates. Therefore force is rate of energy
change regarding distance. If, however, the curve/ be drawn to
a time base as at D, summation will give momentum, or (W#/£•),
for momentum = impulse or ft. Then, for a second definition,
force is rate of change of momentum regarding time. These
principles may be carried much further: thus, kinetic energy is
the sum of \v regarding momentum and so on.

Speaking next of curve averages, it will be easily seen that
average velocity, for example, will depend on the base units, and
that a time-base average can only equal a distance-base average in
the case of uniform velocity, for then distances are proportional
to times, and the curves are exactly alike (see E). Uniformly
accelerated velocity is shewn at F for both time and distance
bases, and the average v is evidently less in the former than
the latter. In like manner there are time- and distance-base
averages of force, as at D and c respectively, a steam-engine-
indicator mean pressure being a distance average. {See p. 1099.)

P. 478. Energy Curves.

Example 63.—A steam piston of a horizontal engine is 30 ins.
diameter, and the net propelling forces due to the steam in Ibs. per
sq. in., at 6 in. intervals of stroke, are as follows : 100, 98, 97, 96, 95,
79, 67, 57, 49, 43, 3*, 34i, 3*, 28, 25, 23, 21, £c. The resistances of
load and friction are 19 tons, assumed constant, and the weight of the
moving parts is taken at 6 tons. Find the correct length of stroke and
draw the velocity curve. (Hons. Applied Mechs. Exam., 1897, slightly

Draw any horizontal base A B, Fig. 824, and set out the 6 in.
spaces as shewn. Erect verticals, plot total net steam pressure in
ton's =p (piston area--2240) = "315^, and draw hyperbolic curve
C Q r>. As the piston advances from left to right, the area under this
curve shews, total energy up to a given position. Summate then
curve C Q D, and obtain curve A M F of total energy (see p. 860). The
scale for B F is obtained from previous explanations. Draw G H so that
AG = P the total load resistance, = 19 tons. Now, energy absorbed
by load is P x D, where D is the stroke swept out up to any point,
and the summation of G H, with same pole P, gives A L j, the curve
of load energy measurable to same scale as B F. But generally—





-,             fciSWE \

Appendix II.                              863

Total energy = Potential energy + Kinetic energy,     and


Total steam energy = (P x D) + w~

where w = weight of moving parts. Hence, as ordinates A M F B
shew total energy, and A L j B the load energy absorbed, the re-
maining ordinates, by deduction, viz., those of A M K L, will indicate
kinetic energy of moving parts. Now,

*2L = K.E.    /. v = V^y^-   X/K.KX4-I4
and V =  VK.E. x 24*8

where energy is in foot tons, and w = 6 tons.     Next, ordinates
K.E. are measured on A L K, and velocity curve found by calculation
The stroke will finish at P, directly under K, where v equals o ; and
the maximum velocity is at N, vertically over Q, where load and steam
curves cross.    Cut off occurs at •£% of stroke A P.

P. 492. Acceleration Curves.—It was shewn at p. 492 how
to construct an acceleration curve to a distance base. The proof
will here be given by reference to Fig. 825. Let V be a velocity
ordinate, whose growth v in a small portion of time / takes place
at A; also let a small distance d be traversed during time /.
x T is a tangent to the curve, and A B a normal; then D B or x
will shew the acceleration fy for


J     t      d    /

But - = velocity V,       for space = tv

and ^ = ™. by similar triangles.    Substituting

The construction cannot be reversed to find v, but \v^ may be
found by summation^ and V be therefrom deduced.

P. 496. Comparison of Angular Velocities in Link-

work. — In pp. 490-496 are found the linear velocities of points in
linkwork. But it is often convenient to know the ratios of
angular velocities in a pair of links, and two cases will here


Appendix IL

be investigated, as examples of slider-crank  and quadriocrank
chains respectively.

Fig. 826 is the linkwork of an oscillating engine, where CP is
the crank and BP the piston rod, and the ratio w of CP to wj. of
B P is required. Linear velocity of P, normally to c P = #, and
normally to B P = v^ Now if i) = s P, v-^=s T, and


^ = — =

x           T5T>



But by similar triangles


sr   v

B P   S P


BP    #
—. —



Also draw-

d v = wj • o P

z> = w x rad. = a) • c P

6U - CP = o^ -OP

B p to F, and draw G F at right angles to B F.
G F || c P, making triangles B F G, B p c similar.

>      OP     BP     B c

~ = — = — = —

>!       C P       P F        C G

or, if B c be angular velocity of crank, c G- is that of piston rod
Then, if c G be turned round to g, a point is found in a polar
curve of angular velocity of B p when w is constant.

In the quadric crank, Fig. 827, let o be the virtual centre.
Then w. A P = zi, and o^ B Q = \. Draw AL || B Q, making
triangles o P Q and A P L similar. Also produce Q P to s, making
triangles SAL and SBQ similar.


o> • AP




1 OQ

A P • AL

_ AP

" AI,



If the chain be that of a beam engine, w is constant and
represented by B Q, while ^ is shewn by A L. Turn A L round on
to line A p and a point I is found in a polar curve which shews
uj-i for any position of crank A p.

P.502. Efficiency of Transruissian by Shafting,— The

Appendix II.


work lost in shafting varies greatly, the limits being 25 to 50%
of the power given; but one should distinguish between instan-
taneous efficiency and that averaged over, say, a whole day.

An extreme case of loss is recorded in Fig. 828, which shews
the H.P. curve of an American workshop during one day.

Neglecting engine friction,

Utilised.          Lost.

Best results for shafting              =     43%            57%

Average results                           =     36%            64%

The highest H.P. given to shafting and machines was 44, and
the average 39. The machines were next driven by electric
motors, with the result shewn, the highest H.P. given to dynamos,

'. 82,8.

motors, and machines being reduced to 29 and the average to 21.
Neglecting the engine loss, as before,

Utilised.           Lost.

Best results for electric driving =     64%           36%

Average results                          =     62%           38%


Appendix IL

The advantage of the alteration is self-evident, being largely
due to the fact that with electric driving the principal losses
vary with the load, and are not constant as with shafting.

P. 504. Velocity-ratio in Hooke's Joint.—Given a
universal or Hooke's joint, its construction is essentially that of
two forks pivoted to a cross, as in Fig. 829. Now let circle
PTQ shew plane of motion of P and Q, while PRQ (an ellipse
in elevation, but really a circle) indicates that of s and R. Let
p move to P! while R moves to Ra; then angles P o Pa and R o Ra

are equal, the latter being that apparent in elevation and not
the real one on plane s R. Draw Ra R2 || R T, and join o R2.
TOR2 is the real angular motion of R, while that of P is POPr
Calling p's velocity v, and R'S velocity ^:

£>!       TR2       RRj       Or

When P arrives at T, and R at Q, the velocity relations are
exactly reversed, for the fork positions have simply interchanged.
Join R Q and draw T w || R Q, thus making o R : O T : : o Q : o w ;
and complete the quarter ellipse R w. Then the, radii vector o P

Appendix II.                             867

to o T shew angular velocity of driver A B, taken constant, while
OR to ow shew varying angular velocity of follower CD. For,
work done being equal, area OPT must equal area o R w, which is
only true for circle and ellipse respectively. Thus, area of circle
= ?r (o x)2, and area of ellipse = TT (o R) (o w), which is TT (ox)2 by

P. 577. Rolling Curves. — Other curves may be produced
in the manner described at p. 517. Thus, when a parabola is
rolled on a straight line, its focus describes a catenary^ the curve
in which a chain hangs, and whose equation is

where e= 2718, and m is a constant depending on the depth of
hang. The ordinate is y and the abscissa x. The tractrix or
anti-friction curve, is the involute of the catenary.

P. 528. Tension of Belts.— Let the arc 0, Fig. 830, of

the pulley A, be wrapped by a belt, the greatest pull Tn being
balanced by /n+ friction. Considering a small elemental strip
On, the tensions /i and / are balanced by pressure /, and /i= /
c diaram

approximately.    By force diagram

p     chord

-             ,  ,      , n

__     on       and/ = / 0n
t     radius

Frictional resistance on small arc = /
/. //u = p. /0n



Appendix IL

which is the small increase of tension for arc 0n.    Hence we may

dt = pie*       and — = p 0n

Summating each side of the equation separately, over the whole

Bat if


.}    _.   Tn=

and Log io

P. 555. Friction. — Experiments on  solid  friction may be
classified under six heads : —

4.  Journal friction.

5.  Collar friction.

6.  Pivot friction.

1.  Static friction.

2.  Friction at low speeds.

3.  Friction at high speeds.

And in every case the surfaces may be wholly or partially
lubricated, may be dry, or may be coated with a resistant. The
first three are usually taken with flat surfaces, and the results
known as (flat friction/ unit pressure being constant over the
surface. In the remaining cases the surface pressure may be
very unequal.

The friction at starting, called friction of repose or static
friction, has been but little examined, but some good results
were obtained in 1895 by Mr. Broomall, of America, being the
averages of twenty or thirty trials with each load.



Cast iron on cast iron    ...

Steel on steel      .........

Steel on cast iron

Cast iron on tin ...
	'454 T

Steel on tin

Cast iron on pine           ...     ....

Pine on pine

Appendix II.


The extreme cases were from -96 to 1-04 times these numbers,
but at certain loads the values became constant As p increased
IUL decreased, except with dry pine on pine, when the reverse
occurred. The effect of long contact was to slightly increase p
with the dry experiments, and markedly so with wet pine on

For low-speed friction, Morin's results may be accepted.
The method of experiment is to find a load P which will just
keep a weight W moving at a slow uniform speed on a level



surface, Fig.  831:   then the ratio P:W = ^, and very regular
results are obtainable.

At high speed somewhat conflicting figures are found.
Apparently there is a decrease in friction with increase of speed
if the surfaces are dry (p. 556), or rather, at high speeds the
friction seems to approach the lubricated cases, most probably
due to a cushion of air drawn in between the surfaces. High-
speed lubricated experiments are troublesome, because the

870                           Appendix II.

lubricant alters with the temperature, every oil having a best
condition, but the results of flat friction at linear velocities from
400 to 1600 ft. per m, shew a constant /* of '23 at 50 Ibs. per
sq. in. ; for cast-iron. surfaces. The apparatus is shewn in Fig. 831,
where A is a rotating disc, and B B two rubbing surfaces (shewn
in detail on the right) pressed upon it by the load c. The
turning effort of D is resisted by string E, and Fn measured by
spring-balance or scale-pan weights.

Beauchamp Tower's journal experiments for the Inst. of
Mech. E. were made at high speeds with heavy loads, and the
projected area Id adopted for measurement of unit pressures,
d being width of journal embraced, and / its length. The
nominal f was thus conventional, the real p often rising to twice
its amount. The load was carried on a knife-edge K, Fig. 832
and c K would assume some position c L. Then

Moment of friction = Moment of weight

The movement was multiplied by levers A and B, but still
being small, was measured by weights in a scale-pan D sufficient
to bring B back to zero. The starting stickiness was overcome
by a small weight E, -and the lever thus kept horizontal for zero
load, the amount being decided by revolving in each direction
and taking the average. Briefly, total friction was found con-
stant at all loads, and p. varied as i ~ P. Oil-bath lubrication
proved best, sponge lubrication some four times as resistant,
and siphons gave still worse results. Temperature materially
altered the friction; thus by lowering from 300° to 120° Fahr.
the friction with lard oil was as 1:3. By means of pressure
gauges it was shewn to be useless to introduce the lubricant
where pressure was greatest, for the latter often rose to 200 Ibs.
per sq, inch and forced out the oil. Oil should therefore enter
at the top of shafts and the bottom of axles, and grooves be cut
to assist it. (See Third Prefaced)

With collar bearings (Inst. Mech. E. experiments) the friction
varied more nearly with the pressure, the apparatus being shewn
in Fig. 832*7. A ring c supported on rollers jj, was pressed

Appendix II.

between discs A and B by the force of a screwed-up spring H, the
amount of pressure being measured by pointer H, and the bolt F
having a spherical seat. Both discs being connected by keys K K,

are revolved by belting, and the friction between the two surfaces
is measured by weight E, in the manner of a dynamometer (see
also p. 558). In marine practice pressures of 50 Ibs. per sq. in.
are allowed on thrust-block collars, and the H.P, transmitted to

872                            Appendix II.

I    "'

;*                              the screw taken at f- the I.H.P.    The direct thrust may be found

from speed, for

V of ship = knots x 101*3

Thrust x knots x 101-3 = effective H.P. x 33000
I                               Qr p = (eff.H.P.)x 33000 ^LHjP,

K x Tor3                     K

- „         f            .    ,     217 I.H.P            I.H.P.

.-. Collar surface required =-------~— = 4*34 ———

50 JtL                         jL

It must be noted that these surfaces' are now horseshoe in
form (see p. 691) and more collars are required than if circular.

For pivot bearings the experimental apparatus in Fig. 833
was adopted, where B is the pivot or footstep fed by oil entering
at pipe H. The shaft D being rotated through bevel gearing A A,
the frictional moment was measured by weight G, which, acting
on pulley F, prevented the bearing at B from rotating. At the
same time, the load was obtained by oil pumped against surfaces
D and E, its intensity being measured by pressure gauge c.

P. 568. Balls arid Live Rollers.—It is found that with
ball or roller bearings the frictional loss is £ or $ of that of a
plain journal, and in large bearings the rollers are kept apart by.
rings, as in Fig. 5 84.

P.575. Efficiencies of 'Machines.—The example on
pp. 571 to 575 shews the methods usually followed in mecha-
nical laboratories to find frictional loss in machines. Two
further cases may be given by way of illustration. Fig. 834 is
a chart of experiments on rope-pulley blocks, a simple fixed
pulley being called a ; i : i system, one movable and one fixed
block a 2:1 system, and so on. Thus 3 upper and 3 lower
pulleys make a 6: i system. Plotting P : W the inclined lines
are obtained, and efficiency curves are further calculated for each
load, A curve of maximum efficiency is also drawn belgw, on a
base of velocity-ratio, which usefully indicates the rapid loss with
increased theoretical advantage. The sheaves were 2§" diameter
and the rope J".
•*                Investigating, we have in a i: i system :

P = W(r +m)       where m is a proper fraction.

Appendix II.

Then, in a 3 : i system, say :
PI= P2(i + «) ...... P2= P3

Adding each side"*:

or     Pa -   ^ (P2 + P3 4- P4)
But since all the rope tensions together = W



and    Pj =  P4+wW

»O          i<3         3O         4-O         So        €<£>         70         80         9.

Again, multiplying sides together in (i);
P4 (i



Substituting this in (2



/         ^3

/        \      V

or    P1(i-,7_i_     =
H       (r+»0v

for a 3 ; i system......(4)

or generally


*  Plf P2, P3 are the rope parts taken in order from P to W.    See Tig. 4.39,


Appendix II.

where the exponent n represents the number of cords; and the
real advantage is thus shewn. The assumption, however, that
pulley friction and rope-bending resistance vary similarly is not
quite true, the latter increasing more rapidly than the former.
Comparing with experiment,

P- W
i : i block;    m =    ——-    =   -195    at max. effy

U1    ,      W
2 : i block ;    — =
'    P


= i*54

which, by experiment =1-52      at max. effy.



= 2*62

4 : i block;    ~

i -


which, by experiment = 2*25      at max. effy.
In the case of a screw jack, the P : W line passed through the


t r*


Fuj. 635.

origin, P & W vanishing simultaneously, because the screw was
reckoned as load. The efficiency was therefore constant at all
loads, being -3535 with a screw 1-92" mean diameter and pitch
£", the velocity ratio being 59-33 : i. The lever was repre-
sented by a pulley mounted on the screw axis. Here again the
efficiency can be found by calculation. Let F G, Fig. 835, be the
mean screw circumference, and H G the pitch, the nut being
moved by force P. Neglecting friction, there are three balancing
forces, P, R, and W; and the force diagram is A B c.







Appendix II.


With friction, the real resistance is R2, and the force diagram
A B D, where 0 = friction angle.

W ^ A B _         i    '

Po         BD

tan (0 + 0)
P          tan 9

P2      tan (0 + 0)

shewing that large pitch is economical.    Returning to the experi-

and efficiency ==


/n      x       tan 0
tan (0 + 0)  = -_- =
effy. "•

•. 0 = (0 + 0)~ 0 = 13*27°-476° = 8-51°

and   IJL — tan 0 = "1494       (Seep. 1125.)

Absorption Dynamometer.— The apparatus in
Fig. 596 is called an Appold 'brake. If P = pull on stud D,
r = radius of brake wheel, and Fn total friction on "brake strap,
the sum of moments being necessarily zero,

(W-S)R ± (PxAD) = Fnr

or the total moment exerted by the-.engine. P may be measured
by two spring balances, one on each side of D, a pull on the right
balance being plus, and on the left minus. The work absorbed
per revolution would be          • ' '

2 ^{(w - S)R ± (P x A D)}

whence B.H.P. is found. If the H.P. be under 15 and the
lubrication sparing, there is little pull on r>, but the lever is
generally a bad arrangement, and a simple strap is now advised,
where only W and S are measured.

P. 580. Distribution of Power.^—We may distinguish
between mere transmission, and distribution from a central station
to many consumers. Professor Unwin has shewn the advantages
of the latter over individual installations, and gives the following

i. Indefinite subdivision and measurement.

af. Minimum first cost and mntxing- loss.           '                    *

3. Freedom from danger.                                         • • . i

3 -M

876                            Appendix II

4.  Consumer's motor to be simple and efficient.

5.  Facility for adaptation to numerous uses.

Only two great natural sources of power have been much used
up to the present, gravity and heat, the former being the water
power of streams, and the latter the latent energy of coal or
petroleum. The distributors have been electricity, wire rope,
compressed air, hydraulics, or coal gas ; of which probably com-
pressed air meets the above requirements most perfectly. Inter-
mittent use means the necessity for storage if a part of the plant
is not to lie idle at times, and electric accumulators are very
costly for large stations, but hydraulic power may be efficiently
stored. A low-pressure water power is used in Switzerland, drawn
from reservoirs 400 ft. high kept filled by turbines, and the
storage of heat has been accomplished by admitting surplus
boiler steam to a large vessel of water under high pressure, the
liquid thus heated being converted into steam when needed, at a
small reduction of pressure.


P. s#j. Conduction of Heat through Plates. — The
boiler furnace having a temperature of about 1500°, and the steam
of 300°, it may be taken that there is a (skin' drop of 500° on
each side, and another drop of 300° in passing through the plate.

P. j#f. Le Chatelier's Pyrometer is undoubtedly the
finest modern apparatus for measuring high temperatures, and, as
improved by Roberts-Austen, is shewn in Fig. 836. A is a thermo-
electric couple of two wires, twisted together, one of platinum and
the other of platinum alloyed with 10% of rhodium. Being pro-
tected with refractory material, they are placed in a source of heat,
and connected to the mirror galvanometer B, within the camera
c. A ray of light from an oxy-hydrogen burner D is deflected and
I N                         focussed on to the mirror E, from which it returns to a photo-

11|                         graphic plate F.    Now the horizontal movement of the light ray at

ff will indicate temperature: but a better way is to cause the plate
F to move vertically by clockwork while the heat changes occur at

Appendix II.


A,  thus  recording a time-temperature curve.    The temperature
scale is graduated by reference to well-known phenomena, whose

FJUV. 836.

temperatures have been accurately obtained by the methods on
p. 587, and the junction at A should directly touch the source of

P. $pj. Steam Drier  or   Separator.—This apparatus is
now considered essential to an engine steam-pipe, especially if the


length of the latter be great, and the principles embodied in
various forms may be illustrated by Fig. 837. Steam enters at A,
and in passing round the helix to escape at c, deposits all the
moisture on the casing wall by centrifugal action, only dry steam
passing away.


Appendix II.

P. 594. Dryness of Steam. — The finding of dryness
value from indicator diagram has already been shewn at p. 764.
Professor Carpenter plots a further curve beneath the expansion
line to represent the fractions thus obtained, and calls this the
1 quality' curve of the steam. It appears that condensation con-
tinues after cut-off A, Fig. 838, till re-evaporation is reached at B,

. 838.

while the final dryness c is often higher than A. To measure the
condensation due to cylinder walls we must know the original
condition of the steam. Several methods are used, but Carpen-
ter's apparatus, Fig. 839, is probably simplest, while being very

accurate. Steam passes from the pipe A into a jacketed separator
B, where all the moisture is deposited. Continuing to the Avater
vessel c, the dry steam is there condensed, and the graduations

Appendix II.                            879

at E and D will respectively shew weights of dry . steam  and
entrained water.

P. 605.   Problems on Energy Changes in a "Working

Gas.—When a gas is expanded or compressed in any manner
behind a working piston,

Heat supplied = change of internal energy + external work

and any of the three terms may be plus, minus, or zero. Thus,
when expanding isothermally, the internal change is zero, and the
gas must be given a positive heat equal to the total work done
(p. 605); and in isothermal compression the heat supply is -
negative, that is, must be abstracted to the same amount as

Example 64. Assuming the p v curve to be hyperbolic, let steam
enter the cylinder at 120 Ibs. absolute pressure per square in., tem-
perature 348° F., and expand to 30 Ibs. absolute, temperature 250° F.
Find heat supplied per Ib. weight. (Hons. Steam Exam. 1894.)

Spec. vol. at 250° = 13*5 cub. ft.       and P = (144x30) Ibs.
.-.    p V = 144x30 x 13*5 = 58320 ft. Ibs. ; which is constant.


= 4        and loger =. 1*3863

P        30

Total heat at first,   Hj = S^Lj = 348-32 + 871 = 1187 B.T.U.

Total heat at end,    H2 = S2H-L2 = 250-32 + 940 =1158 B,T.U.

Loss of internal energy = 772 (Hj, - P V) - 772 (H2 - P V)

= 22388 ft. Ibs.
.".    Heat supplied = external work - internal change

= PVloge7-~ 22388    =    (58320X1*3863)-22388

= 58461 ft. Ibs., or 747 B.T.U.   per 13^ cub. it, or per I Ib. weight.

[N.B.—The real result should be 87 B.T.U., the discrepancy being
due to assuming the expansion of dry ste